Model results

ARC-AGI

Choose a shared evaluation protocol to compare model scores. Versions, sources and configurations stay separate.

Atlas snapshot captured 10/19/2026, 10:20:28 AM UTC. This is an import timestamp, not the evaluation date. Coverage may be incomplete.

Reviewed recommendation and limits

A useful test of unfamiliar-rule learning, with published testing policy and configurations.

ARC-AGI-1, 2 and 3 are distinct protocols. Imported groups below may use different protocols; this review does not certify each run. A strong ARC result does not establish workplace competence. Keep standard and provider-adapter runs apart.

Publisher results ARC Prize testing policy

Evaluation groups

Open a group to see its exact protocol and scores. Different configurations remain separate.

No model rankings are available in this snapshot. Reported results remain available below.

Other reported results

These reports have no model ranking. Open a group for its evidence and the reason positions are withheld.

Claude Opus 4.7v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Opus 4.7 (High); CoT; evaluation cost $10000; ; result date not supplied; so…
ARC-AGI-3 semi-private; Opus 4.7 (High); CoT; evaluation cost $10000; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.…
ARC-AGI-3 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.; evaluation cost $10000; ; result date not supplied; source captured 2026-09-10
Claude Opus 5v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; …
ARC-AGI-3 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; evaluation cost $20657.37; result details https://arcprize.org/results/anthropic-claude-opus-5; result date not supplied; source captured 2026-09-10
GPT-5.4v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.4 (High); CoT; evaluation cost $5187.41; ; result date not supplied; s…
ARC-AGI-3 semi-private; GPT-5.4 (High); CoT; evaluation cost $5187.41; ; result date not supplied; source captured 2026-09-10
GPT-5.5v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.5 (High); CoT; evaluation cost $10000; ; result date not supplied; sou…
ARC-AGI-3 semi-private; GPT-5.5 (High); CoT; evaluation cost $10000; ; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Luna (High); CoT; evaluation cost $2482.44; result details https://a…
ARC-AGI-3 semi-private; GPT-5.6 Luna (High); CoT; evaluation cost $2482.44; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Luna (Low); CoT; evaluation cost $2307.49; result details https://ar…
ARC-AGI-3 semi-private; GPT-5.6 Luna (Low); CoT; evaluation cost $2307.49; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Luna (Max); CoT; evaluation cost $3189.42; result details https://ar…
ARC-AGI-3 semi-private; GPT-5.6 Luna (Max); CoT; evaluation cost $3189.42; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Luna (Medium); CoT; evaluation cost $2379.13; result details https:/…
ARC-AGI-3 semi-private; GPT-5.6 Luna (Medium); CoT; evaluation cost $2379.13; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Luna (XHigh); CoT; evaluation cost $2801.59; result details https://…
ARC-AGI-3 semi-private; GPT-5.6 Luna (XHigh); CoT; evaluation cost $2801.59; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Sol (High); CoT; evaluation cost $15176.07; result details https://a…
ARC-AGI-3 semi-private; GPT-5.6 Sol (High); CoT; evaluation cost $15176.07; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Sol (Low); CoT; evaluation cost $12805.6; result details https://arc…
ARC-AGI-3 semi-private; GPT-5.6 Sol (Low); CoT; evaluation cost $12805.6; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Sol (Max); CoT; evaluation cost $25064.11; result details https://ar…
ARC-AGI-3 semi-private; GPT-5.6 Sol (Max); CoT; evaluation cost $25064.11; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Sol (Medium); CoT; evaluation cost $12971.17; result details https:/…
ARC-AGI-3 semi-private; GPT-5.6 Sol (Medium); CoT; evaluation cost $12971.17; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Sol (XHigh); CoT; evaluation cost $19216.38; result details https://…
ARC-AGI-3 semi-private; GPT-5.6 Sol (XHigh); CoT; evaluation cost $19216.38; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Terra (High); CoT; evaluation cost $5881.39; result details https://…
ARC-AGI-3 semi-private; GPT-5.6 Terra (High); CoT; evaluation cost $5881.39; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Terra (Low); CoT; evaluation cost $5068.48; result details https://a…
ARC-AGI-3 semi-private; GPT-5.6 Terra (Low); CoT; evaluation cost $5068.48; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Terra (Max); CoT; evaluation cost $7918.31; result details https://a…
ARC-AGI-3 semi-private; GPT-5.6 Terra (Max); CoT; evaluation cost $7918.31; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Terra (Medium); CoT; evaluation cost $5895.31; result details https:…
ARC-AGI-3 semi-private; GPT-5.6 Terra (Medium); CoT; evaluation cost $5895.31; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-5.6 Terra (XHigh); CoT; evaluation cost $6804.04; result details https:/…
ARC-AGI-3 semi-private; GPT-5.6 Terra (XHigh); CoT; evaluation cost $6804.04; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (High); CoT; Evaluation conducted with 'high'…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (High); CoT; Evaluation conducted with 'high' reasoning effort using the provider adapter harness.; evaluation cost $18816.63138; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Low); CoT; Evaluation conducted with 'low' r…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Low); CoT; Evaluation conducted with 'low' reasoning effort using the provider adapter harness.; evaluation cost $21297.922310000005; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Max); CoT; Evaluation conducted with 'max' r…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Max); CoT; Evaluation conducted with 'max' reasoning effort using the provider adapter harness.; evaluation cost $17331.95327; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Medium); CoT; Evaluation conducted with 'med…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (Medium); CoT; Evaluation conducted with 'medium' reasoning effort using the provider adapter harness.; evaluation cost $19284.776630000004; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (None); CoT; Evaluation conducted with 'none'…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (None); CoT; Evaluation conducted with 'none' reasoning effort using the provider adapter harness.; evaluation cost $23456.663940000002; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (XHigh); CoT; Evaluation conducted with 'xhig…
ARC-AGI-3 semi-private; GPT-6 Astra - Provider Adapter (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort using the provider adapter harness.; evaluation cost $18147.191119999996; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.;…
ARC-AGI-3 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.; evaluation cost $40704.722830000006; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; e…
ARC-AGI-3 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; evaluation cost $38166.47030000001; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; e…
ARC-AGI-3 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; evaluation cost $26097.501720000007; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effo…
ARC-AGI-3 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; evaluation cost $48090.34598999999; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.;…
ARC-AGI-3 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.; evaluation cost $49791.11317000001; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort…
ARC-AGI-3 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; evaluation cost $37317.38768; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
Grok 4.5v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Grok 4.5 (High); CoT; evaluation cost $6892.87; result details https://arcpr…
ARC-AGI-3 semi-private; Grok 4.5 (High); CoT; evaluation cost $6892.87; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Grok 4.5 (Low); CoT; evaluation cost $8718.48; result details https://arcpri…
ARC-AGI-3 semi-private; Grok 4.5 (Low); CoT; evaluation cost $8718.48; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Grok 4.5 (Medium); CoT; evaluation cost $8458.3; result details https://arcp…
ARC-AGI-3 semi-private; Grok 4.5 (Medium); CoT; evaluation cost $8458.3; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.6v3_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-3 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; …
ARC-AGI-3 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; evaluation cost $5612; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Claude Fable 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5 (High); CoT; Evaluation conducted with 'high' reasoning effor…
ARC-AGI-2 semi-private; Claude Fable 5 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $2.4358764166666664; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5 (Low); CoT; Evaluation conducted with 'low' reasoning effort.…
ARC-AGI-2 semi-private; Claude Fable 5 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $1.4489280833333338; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5 (Max); CoT; Evaluation conducted with 'max' reasoning effort.…
ARC-AGI-2 semi-private; Claude Fable 5 (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $5.449076000000001; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5 (Medium); CoT; Evaluation conducted with 'medium' reasoning e…
ARC-AGI-2 semi-private; Claude Fable 5 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $1.9269780833333334; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning eff…
ARC-AGI-2 semi-private; Claude Fable 5 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $3.0043564166666665; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5.1 (High); CoT; Evaluation conducted with 'high' reasoning eff…
ARC-AGI-2 semi-private; Claude Fable 5.1 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $1.6673061666666666; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5.1 (Low); CoT; Evaluation conducted with 'low' reasoning effor…
ARC-AGI-2 semi-private; Claude Fable 5.1 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.9523265833333336; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5.1 (Max); CoT; Evaluation conducted with 'max' reasoning effor…
ARC-AGI-2 semi-private; Claude Fable 5.1 (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $4.492348833333334; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5.1 (Medium); CoT; Evaluation conducted with 'medium' reasoning…
ARC-AGI-2 semi-private; Claude Fable 5.1 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $1.2239411666666673; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Fable 5.1 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning e…
ARC-AGI-2 semi-private; Claude Fable 5.1 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $3.120216916666669; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 1K); CoT; cost per task $0.0471; ; result date no…
ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 1K); CoT; cost per task $0.0471; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 8K); CoT; cost per task $0.0909; ; result date no…
ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 8K); CoT; cost per task $0.0909; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 16K); CoT; cost per task $0.1392; ; result date n…
ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 16K); CoT; cost per task $0.1392; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 32K); CoT; cost per task $0.3766; ; result date n…
ARC-AGI-2 semi-private; Claude Haiku 4.5 (Thinking 32K); CoT; cost per task $0.3766; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Haiku 4.5; Base LLM; cost per task $0.0426; ; result date not supplie…
ARC-AGI-2 semi-private; Claude Haiku 4.5; Base LLM; cost per task $0.0426; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 1K); CoT; cost per task $0.7503; ; result date not s…
ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 1K); CoT; cost per task $0.7503; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 8K); CoT; cost per task $1.1569; ; result date not s…
ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 8K); CoT; cost per task $1.1569; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 16K); CoT; cost per task $1.9284; ; result date not …
ARC-AGI-2 semi-private; Claude Opus 4 (Thinking 16K); CoT; cost per task $1.9284; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 8K); CoT; cost per task $0.48; ; result date not supplie…
ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 8K); CoT; cost per task $0.48; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 16K); CoT; cost per task $0.79; ; result date not suppli…
ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 16K); CoT; cost per task $0.79; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 64K); CoT; cost per task $2.4; ; result date not supplie…
ARC-AGI-2 semi-private; Opus 4.5 (Thinking, 64K); CoT; cost per task $2.4; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Opus 4.5 (Thinking, None); CoT; cost per task $0.22; ; result date not suppl…
ARC-AGI-2 semi-private; Opus 4.5 (Thinking, None); CoT; cost per task $0.22; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, High); CoT; Evaluation conducted with a 120K thinking…
ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, High); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'high' output effort.; cost per task $3.4719; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Low); CoT; Evaluation conducted with a 120K thinking …
ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Low); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'low' output effort.; cost per task $2.2507; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Max); CoT; Evaluation conducted with a 120K thinking …
ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Max); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'max' output effort.; cost per task $3.641; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Medium); CoT; Evaluation conducted with a 120K thinki…
ARC-AGI-2 semi-private; Claude Opus 4.6 (120K, Medium); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'medium' output effort.; cost per task $2.7209; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude 4.7 (High); CoT; cost per task $3.17; ; result date not supplied; sou…
ARC-AGI-2 semi-private; Claude 4.7 (High); CoT; cost per task $3.17; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude 4.7 (Low); CoT; cost per task $2.38; ; result date not supplied; sour…
ARC-AGI-2 semi-private; Claude 4.7 (Low); CoT; cost per task $2.38; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude 4.7 (Max); CoT; cost per task $7.43; ; result date not supplied; sour…
ARC-AGI-2 semi-private; Claude 4.7 (Max); CoT; cost per task $7.43; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude 4.7 (Medium); CoT; cost per task $2.96; ; result date not supplied; s…
ARC-AGI-2 semi-private; Claude 4.7 (Medium); CoT; cost per task $2.96; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.…
ARC-AGI-2 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.; cost per task $2.7425; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.8 (Low); CoT; Evaluation conducted with 'low' output effort.; …
ARC-AGI-2 semi-private; Claude Opus 4.8 (Low); CoT; Evaluation conducted with 'low' output effort.; cost per task $1.6818; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 4.8 (Medium); CoT; Evaluation conducted with 'medium' output eff…
ARC-AGI-2 semi-private; Claude Opus 4.8 (Medium); CoT; Evaluation conducted with 'medium' output effort.; cost per task $2.3896; ; result date not supplied; source captured 2026-09-10
Claude Opus 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; …
ARC-AGI-2 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; cost per task $1.45; result details https://arcprize.org/results/anthropic-claude-opus-5; result date not supplied; source captured 2026-09-10
Claude Opus 5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Opus 5 (Max); CoT; Evaluation conducted with 'max' output effort.; co…
ARC-AGI-2 semi-private; Claude Opus 5 (Max); CoT; Evaluation conducted with 'max' output effort.; cost per task $2.06; result details https://arcprize.org/results/anthropic-claude-opus-5; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 1K); CoT; cost per task $0.1425; ; result date not…
ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 1K); CoT; cost per task $0.1425; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 8K); CoT; cost per task $0.2654; ; result date not…
ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 8K); CoT; cost per task $0.2654; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 16K); CoT; cost per task $0.4857; ; result date no…
ARC-AGI-2 semi-private; Claude Sonnet 4 (Thinking 16K); CoT; cost per task $0.4857; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 1K); CoT; cost per task $0.1424; ; result date n…
ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 1K); CoT; cost per task $0.1424; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 8K); CoT; cost per task $0.2349; ; result date n…
ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 8K); CoT; cost per task $0.2349; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 16K); CoT; cost per task $0.3499; ; result date …
ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 16K); CoT; cost per task $0.3499; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 32K); CoT; cost per task $0.7589; ; result date …
ARC-AGI-2 semi-private; Claude Sonnet 4.5 (Thinking 32K); CoT; cost per task $0.7589; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.5; Base LLM; cost per task $0.1295; ; result date not suppli…
ARC-AGI-2 semi-private; Claude Sonnet 4.5; Base LLM; cost per task $0.1295; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.6 (High); CoT; Evaluation conducted with a 120K thinking bud…
ARC-AGI-2 semi-private; Claude Sonnet 4.6 (High); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'high' output effort.; cost per task $2.6966; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Claude Sonnet 4.6 (Max); CoT; Evaluation conducted with a 120K thinking budg…
ARC-AGI-2 semi-private; Claude Sonnet 4.6 (Max); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'max' output effort.; cost per task $2.7179; ; result date not supplied; source captured 2026-09-10
GPT-4.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-4.1-Mini; Base LLM; cost per task $0.0139; ; result date not supplied; s…
ARC-AGI-2 semi-private; GPT-4.1-Mini; Base LLM; cost per task $0.0139; ; result date not supplied; source captured 2026-09-10
GPT-4.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-4.1-Nano; Base LLM; cost per task $0.0036; ; result date not supplied; s…
ARC-AGI-2 semi-private; GPT-4.1-Nano; Base LLM; cost per task $0.0036; ; result date not supplied; source captured 2026-09-10
GPT-4.1v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-4.1; Base LLM; cost per task $0.0691; ; result date not supplied; source…
ARC-AGI-2 semi-private; GPT-4.1; Base LLM; cost per task $0.0691; ; result date not supplied; source captured 2026-09-10
GPT-5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5 (High); CoT; cost per task $0.7302; ; result date not supplied; source…
ARC-AGI-2 semi-private; GPT-5 (High); CoT; cost per task $0.7302; ; result date not supplied; source captured 2026-09-10
GPT-5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5 (Low); CoT; cost per task $0.1896; ; result date not supplied; source …
ARC-AGI-2 semi-private; GPT-5 (Low); CoT; cost per task $0.1896; ; result date not supplied; source captured 2026-09-10
GPT-5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5 (Medium); CoT; cost per task $0.4486; ; result date not supplied; sour…
ARC-AGI-2 semi-private; GPT-5 (Medium); CoT; cost per task $0.4486; ; result date not supplied; source captured 2026-09-10
GPT-5.4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.4 (High); CoT; cost per task $1.02; ; result date not supplied; source…
ARC-AGI-2 semi-private; GPT-5.4 (High); CoT; cost per task $1.02; ; result date not supplied; source captured 2026-09-10
GPT-5.4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.4 (Low); CoT; cost per task $0.27; ; result date not supplied; source …
ARC-AGI-2 semi-private; GPT-5.4 (Low); CoT; cost per task $0.27; ; result date not supplied; source captured 2026-09-10
GPT-5.4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.4 (Medium); CoT; cost per task $0.68; ; result date not supplied; sour…
ARC-AGI-2 semi-private; GPT-5.4 (Medium); CoT; cost per task $0.68; ; result date not supplied; source captured 2026-09-10
GPT-5.4v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.4 (XHigh); CoT; cost per task $1.52; ; result date not supplied; sourc…
ARC-AGI-2 semi-private; GPT-5.4 (XHigh); CoT; cost per task $1.52; ; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Luna (High); CoT; cost per task $0.29; result details https://arcpri…
ARC-AGI-2 semi-private; GPT-5.6 Luna (High); CoT; cost per task $0.29; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Luna (Low); CoT; cost per task $0.07; result details https://arcpriz…
ARC-AGI-2 semi-private; GPT-5.6 Luna (Low); CoT; cost per task $0.07; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Luna (Max); CoT; cost per task $0.67; result details https://arcpriz…
ARC-AGI-2 semi-private; GPT-5.6 Luna (Max); CoT; cost per task $0.67; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Luna (Medium); CoT; cost per task $0.11; result details https://arcp…
ARC-AGI-2 semi-private; GPT-5.6 Luna (Medium); CoT; cost per task $0.11; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Luna (XHigh); CoT; cost per task $0.54; result details https://arcpr…
ARC-AGI-2 semi-private; GPT-5.6 Luna (XHigh); CoT; cost per task $0.54; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Sol (High); CoT; cost per task $0.74; result details https://arcpriz…
ARC-AGI-2 semi-private; GPT-5.6 Sol (High); CoT; cost per task $0.74; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Sol (Low); CoT; cost per task $0.32; result details https://arcprize…
ARC-AGI-2 semi-private; GPT-5.6 Sol (Low); CoT; cost per task $0.32; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Sol (Max); CoT; cost per task $1.44; result details https://arcprize…
ARC-AGI-2 semi-private; GPT-5.6 Sol (Max); CoT; cost per task $1.44; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Sol (Medium); CoT; cost per task $0.47; result details https://arcpr…
ARC-AGI-2 semi-private; GPT-5.6 Sol (Medium); CoT; cost per task $0.47; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Sol (XHigh); CoT; cost per task $1.04; result details https://arcpri…
ARC-AGI-2 semi-private; GPT-5.6 Sol (XHigh); CoT; cost per task $1.04; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Terra (High); CoT; cost per task $0.55; result details https://arcpr…
ARC-AGI-2 semi-private; GPT-5.6 Terra (High); CoT; cost per task $0.55; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Terra (Low); CoT; cost per task $0.17; result details https://arcpri…
ARC-AGI-2 semi-private; GPT-5.6 Terra (Low); CoT; cost per task $0.17; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Terra (Max); CoT; cost per task $1.09; result details https://arcpri…
ARC-AGI-2 semi-private; GPT-5.6 Terra (Max); CoT; cost per task $1.09; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Terra (Medium); CoT; cost per task $0.27; result details https://arc…
ARC-AGI-2 semi-private; GPT-5.6 Terra (Medium); CoT; cost per task $0.27; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-5.6 Terra (XHigh); CoT; cost per task $0.69; result details https://arcp…
ARC-AGI-2 semi-private; GPT-5.6 Terra (XHigh); CoT; cost per task $0.69; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.;…
ARC-AGI-2 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.6680793333333332; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; c…
ARC-AGI-2 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.4162418333333334; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; c…
ARC-AGI-2 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $1.1194680833333335; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effo…
ARC-AGI-2 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.4800926666666667; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.;…
ARC-AGI-2 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.; cost per task $0.3696743333333333; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort…
ARC-AGI-2 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $0.8286855833333331; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
Grok 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.5 (High); CoT; cost per task $0.78; result details https://arcprize.o…
ARC-AGI-2 semi-private; Grok 4.5 (High); CoT; cost per task $0.78; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.5 (Low); CoT; cost per task $0.4; result details https://arcprize.org…
ARC-AGI-2 semi-private; Grok 4.5 (Low); CoT; cost per task $0.4; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.5 (Medium); CoT; cost per task $0.75; result details https://arcprize…
ARC-AGI-2 semi-private; Grok 4.5 (Medium); CoT; cost per task $0.75; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.6 (High); CoT; Evaluation conducted with 'high' reasoning effort.; co…
ARC-AGI-2 semi-private; Grok 4.6 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.7338575666666665; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.6 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost…
ARC-AGI-2 semi-private; Grok 4.6 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.2124022166666667; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.6 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.…
ARC-AGI-2 semi-private; Grok 4.6 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.6586968666666668; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; …
ARC-AGI-2 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $0.7574904499999999; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
o3v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3 (High); CoT; cost per task $0.8339; ; result date not supplied; source ca…
ARC-AGI-2 semi-private; o3 (High); CoT; cost per task $0.8339; ; result date not supplied; source captured 2026-09-10
o3v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3 (Low); CoT; cost per task $0.2343; ; result date not supplied; source cap…
ARC-AGI-2 semi-private; o3 (Low); CoT; cost per task $0.2343; ; result date not supplied; source captured 2026-09-10
o3v2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3 (Medium); CoT; cost per task $0.4787; ; result date not supplied; source …
ARC-AGI-2 semi-private; o3 (Medium); CoT; cost per task $0.4787; ; result date not supplied; source captured 2026-09-10
o3 Miniv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-mini (High); CoT; cost per task $0.5472; ; result date not supplied; sour…
ARC-AGI-2 semi-private; o3-mini (High); CoT; cost per task $0.5472; ; result date not supplied; source captured 2026-09-10
o3 Miniv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-mini (Low); CoT; cost per task $0.0623; ; result date not supplied; sourc…
ARC-AGI-2 semi-private; o3-mini (Low); CoT; cost per task $0.0623; ; result date not supplied; source captured 2026-09-10
o3 Miniv2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-mini (Medium); CoT; cost per task $0.2843; ; result date not supplied; so…
ARC-AGI-2 semi-private; o3-mini (Medium); CoT; cost per task $0.2843; ; result date not supplied; source captured 2026-09-10
o3 Prov2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-Pro (High); CoT + Synthesis; cost per task $7.5516; ; result date not sup…
ARC-AGI-2 semi-private; o3-Pro (High); CoT + Synthesis; cost per task $7.5516; ; result date not supplied; source captured 2026-09-10
o3 Prov2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-Pro (Low); CoT + Synthesis; cost per task $2.2293; ; result date not supp…
ARC-AGI-2 semi-private; o3-Pro (Low); CoT + Synthesis; cost per task $2.2293; ; result date not supplied; source captured 2026-09-10
o3 Prov2_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-2 semi-private; o3-Pro (Medium); CoT + Synthesis; cost per task $4.7441; ; result date not s…
ARC-AGI-2 semi-private; o3-Pro (Medium); CoT + Synthesis; cost per task $4.7441; ; result date not supplied; source captured 2026-09-10
Claude Fable 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5 (High); CoT; Evaluation conducted with 'high' reasoning effor…
ARC-AGI-1 semi-private; Claude Fable 5 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.7666246999999999; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5 (Low); CoT; Evaluation conducted with 'low' reasoning effort.…
ARC-AGI-1 semi-private; Claude Fable 5 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.5076897000000001; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5 (Max); CoT; Evaluation conducted with 'max' reasoning effort.…
ARC-AGI-1 semi-private; Claude Fable 5 (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $2.1101577000000007; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5 (Medium); CoT; Evaluation conducted with 'medium' reasoning e…
ARC-AGI-1 semi-private; Claude Fable 5 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.6638002000000002; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning eff…
ARC-AGI-1 semi-private; Claude Fable 5 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $1.0224337000000001; result details https://arcprize.org/results/anthropic-claude-fable-5; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5.1 (High); CoT; Evaluation conducted with 'high' reasoning eff…
ARC-AGI-1 semi-private; Claude Fable 5.1 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.5169491; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5.1 (Low); CoT; Evaluation conducted with 'low' reasoning effor…
ARC-AGI-1 semi-private; Claude Fable 5.1 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.36604360000000014; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5.1 (Max); CoT; Evaluation conducted with 'max' reasoning effor…
ARC-AGI-1 semi-private; Claude Fable 5.1 (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $1.3997611; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5.1 (Medium); CoT; Evaluation conducted with 'medium' reasoning…
ARC-AGI-1 semi-private; Claude Fable 5.1 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.4425576000000001; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Fable 5.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Fable 5.1 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning e…
ARC-AGI-1 semi-private; Claude Fable 5.1 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $0.9578876000000003; result details https://arcprize.org/results/anthropic-claude-fable-5-1; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 1K); CoT; cost per task $0.0306; ; result date no…
ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 1K); CoT; cost per task $0.0306; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 8K); CoT; cost per task $0.0687; ; result date no…
ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 8K); CoT; cost per task $0.0687; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 16K); CoT; cost per task $0.1042; ; result date n…
ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 16K); CoT; cost per task $0.1042; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 32K); CoT; cost per task $0.2558; ; result date n…
ARC-AGI-1 semi-private; Claude Haiku 4.5 (Thinking 32K); CoT; cost per task $0.2558; ; result date not supplied; source captured 2026-09-10
Claude Haiku 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Haiku 4.5; Base LLM; cost per task $0.0263; ; result date not supplie…
ARC-AGI-1 semi-private; Claude Haiku 4.5; Base LLM; cost per task $0.0263; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 1K); CoT; cost per task $0.5021; ; result date not s…
ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 1K); CoT; cost per task $0.5021; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 8K); CoT; cost per task $0.7408; ; result date not s…
ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 8K); CoT; cost per task $0.7408; ; result date not supplied; source captured 2026-09-10
Claude Opus 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 16K); CoT; cost per task $1.2496; ; result date not …
ARC-AGI-1 semi-private; Claude Opus 4 (Thinking 16K); CoT; cost per task $1.2496; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 8K); CoT; cost per task $0.38; ; result date not supplie…
ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 8K); CoT; cost per task $0.38; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 16K); CoT; cost per task $0.62; ; result date not suppli…
ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 16K); CoT; cost per task $0.62; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 32K); CoT; cost per task $0.95; ; result date not suppli…
ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 32K); CoT; cost per task $0.95; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 64K); CoT; cost per task $1.47; ; result date not suppli…
ARC-AGI-1 semi-private; Opus 4.5 (Thinking, 64K); CoT; cost per task $1.47; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Opus 4.5 (Thinking, None); CoT; cost per task $0.13; ; result date not suppl…
ARC-AGI-1 semi-private; Opus 4.5 (Thinking, None); CoT; cost per task $0.13; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, High); CoT; Evaluation conducted with a 120K thinking…
ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, High); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'high' output effort.; cost per task $1.3948; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Low); CoT; Evaluation conducted with a 120K thinking …
ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Low); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'low' output effort.; cost per task $0.648; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Max); CoT; Evaluation conducted with a 120K thinking …
ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Max); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'max' output effort.; cost per task $1.8791; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Medium); CoT; Evaluation conducted with a 120K thinki…
ARC-AGI-1 semi-private; Claude Opus 4.6 (120K, Medium); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'medium' output effort.; cost per task $0.9647; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude 4.7 (High); CoT; cost per task $1.41; ; result date not supplied; sou…
ARC-AGI-1 semi-private; Claude 4.7 (High); CoT; cost per task $1.41; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude 4.7 (Low); CoT; cost per task $0.76; ; result date not supplied; sour…
ARC-AGI-1 semi-private; Claude 4.7 (Low); CoT; cost per task $0.76; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude 4.7 (Max); CoT; cost per task $2.58; ; result date not supplied; sour…
ARC-AGI-1 semi-private; Claude 4.7 (Max); CoT; cost per task $2.58; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.7v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude 4.7 (Medium); CoT; cost per task $1.04; ; result date not supplied; s…
ARC-AGI-1 semi-private; Claude 4.7 (Medium); CoT; cost per task $1.04; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.…
ARC-AGI-1 semi-private; Claude Opus 4.8 (High); CoT; Evaluation conducted with 'high' output effort.; cost per task $1.0416; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.8 (Low); CoT; Evaluation conducted with 'low' output effort.; …
ARC-AGI-1 semi-private; Claude Opus 4.8 (Low); CoT; Evaluation conducted with 'low' output effort.; cost per task $0.6711; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.8 (Max); CoT; Evaluation conducted with 'max' output effort.; …
ARC-AGI-1 semi-private; Claude Opus 4.8 (Max); CoT; Evaluation conducted with 'max' output effort.; cost per task $2.332; ; result date not supplied; source captured 2026-09-10
Claude Opus 4.8v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 4.8 (Medium); CoT; Evaluation conducted with 'medium' output eff…
ARC-AGI-1 semi-private; Claude Opus 4.8 (Medium); CoT; Evaluation conducted with 'medium' output effort.; cost per task $0.9115; ; result date not supplied; source captured 2026-09-10
Claude Opus 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; …
ARC-AGI-1 semi-private; Claude Opus 5 (High); CoT; Evaluation conducted with 'high' output effort.; cost per task $0.47; result details https://arcprize.org/results/anthropic-claude-opus-5; result date not supplied; source captured 2026-09-10
Claude Opus 5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Opus 5 (Max); CoT; Evaluation conducted with 'max' output effort.; co…
ARC-AGI-1 semi-private; Claude Opus 5 (Max); CoT; Evaluation conducted with 'max' output effort.; cost per task $0.7; result details https://arcprize.org/results/anthropic-claude-opus-5; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 1K); CoT; cost per task $0.0937; ; result date not…
ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 1K); CoT; cost per task $0.0937; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 8K); CoT; cost per task $0.1952; ; result date not…
ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 8K); CoT; cost per task $0.1952; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 16K); CoT; cost per task $0.3658; ; result date no…
ARC-AGI-1 semi-private; Claude Sonnet 4 (Thinking 16K); CoT; cost per task $0.3658; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 1K); CoT; cost per task $0.0928; ; result date n…
ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 1K); CoT; cost per task $0.0928; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 8K); CoT; cost per task $0.1776; ; result date n…
ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 8K); CoT; cost per task $0.1776; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 16K); CoT; cost per task $0.2683; ; result date …
ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 16K); CoT; cost per task $0.2683; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 32K); CoT; cost per task $0.5193; ; result date …
ARC-AGI-1 semi-private; Claude Sonnet 4.5 (Thinking 32K); CoT; cost per task $0.5193; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.5; Base LLM; cost per task $0.0808; ; result date not suppli…
ARC-AGI-1 semi-private; Claude Sonnet 4.5; Base LLM; cost per task $0.0808; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.6 (High); CoT; Evaluation conducted with a 120K thinking bud…
ARC-AGI-1 semi-private; Claude Sonnet 4.6 (High); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'high' output effort.; cost per task $1.1908; ; result date not supplied; source captured 2026-09-10
Claude Sonnet 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Claude Sonnet 4.6 (Max); CoT; Evaluation conducted with a 120K thinking budg…
ARC-AGI-1 semi-private; Claude Sonnet 4.6 (Max); CoT; Evaluation conducted with a 120K thinking budget and a maximum context window of 128K tokens and 'max' output effort.; cost per task $1.452; ; result date not supplied; source captured 2026-09-10
GPT-4.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-4.1-Mini; Base LLM; cost per task $0.0078; ; result date not supplied; s…
ARC-AGI-1 semi-private; GPT-4.1-Mini; Base LLM; cost per task $0.0078; ; result date not supplied; source captured 2026-09-10
GPT-4.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-4.1-Nano; Base LLM; cost per task $0.0021; ; result date not supplied; s…
ARC-AGI-1 semi-private; GPT-4.1-Nano; Base LLM; cost per task $0.0021; ; result date not supplied; source captured 2026-09-10
GPT-4.1v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-4.1; Base LLM; cost per task $0.039; ; result date not supplied; source …
ARC-AGI-1 semi-private; GPT-4.1; Base LLM; cost per task $0.039; ; result date not supplied; source captured 2026-09-10
GPT-5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5 (High); CoT; cost per task $0.5087; ; result date not supplied; source…
ARC-AGI-1 semi-private; GPT-5 (High); CoT; cost per task $0.5087; ; result date not supplied; source captured 2026-09-10
GPT-5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5 (Low); CoT; cost per task $0.1531; ; result date not supplied; source …
ARC-AGI-1 semi-private; GPT-5 (Low); CoT; cost per task $0.1531; ; result date not supplied; source captured 2026-09-10
GPT-5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5 (Medium); CoT; cost per task $0.3301; ; result date not supplied; sour…
ARC-AGI-1 semi-private; GPT-5 (Medium); CoT; cost per task $0.3301; ; result date not supplied; source captured 2026-09-10
GPT-5.4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.4 (High); CoT; cost per task $0.37; ; result date not supplied; source…
ARC-AGI-1 semi-private; GPT-5.4 (High); CoT; cost per task $0.37; ; result date not supplied; source captured 2026-09-10
GPT-5.4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.4 (Low); CoT; cost per task $0.15; ; result date not supplied; source …
ARC-AGI-1 semi-private; GPT-5.4 (Low); CoT; cost per task $0.15; ; result date not supplied; source captured 2026-09-10
GPT-5.4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.4 (Medium); CoT; cost per task $0.25; ; result date not supplied; sour…
ARC-AGI-1 semi-private; GPT-5.4 (Medium); CoT; cost per task $0.25; ; result date not supplied; source captured 2026-09-10
GPT-5.4v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.4 (XHigh); CoT; cost per task $0.62; ; result date not supplied; sourc…
ARC-AGI-1 semi-private; GPT-5.4 (XHigh); CoT; cost per task $0.62; ; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Luna (High); CoT; cost per task $0.13; result details https://arcpri…
ARC-AGI-1 semi-private; GPT-5.6 Luna (High); CoT; cost per task $0.13; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Luna (Low); CoT; cost per task $0.04; result details https://arcpriz…
ARC-AGI-1 semi-private; GPT-5.6 Luna (Low); CoT; cost per task $0.04; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Luna (Max); CoT; cost per task $0.32; result details https://arcpriz…
ARC-AGI-1 semi-private; GPT-5.6 Luna (Max); CoT; cost per task $0.32; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Luna (Medium); CoT; cost per task $0.06; result details https://arcp…
ARC-AGI-1 semi-private; GPT-5.6 Luna (Medium); CoT; cost per task $0.06; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Lunav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Luna (XHigh); CoT; cost per task $0.21; result details https://arcpr…
ARC-AGI-1 semi-private; GPT-5.6 Luna (XHigh); CoT; cost per task $0.21; result details https://arcprize.org/results/openai-gpt-5-6-luna; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Sol (High); CoT; cost per task $0.3; result details https://arcprize…
ARC-AGI-1 semi-private; GPT-5.6 Sol (High); CoT; cost per task $0.3; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Sol (Low); CoT; cost per task $0.17; result details https://arcprize…
ARC-AGI-1 semi-private; GPT-5.6 Sol (Low); CoT; cost per task $0.17; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Sol (Max); CoT; cost per task $0.54; result details https://arcprize…
ARC-AGI-1 semi-private; GPT-5.6 Sol (Max); CoT; cost per task $0.54; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Sol (Medium); CoT; cost per task $0.22; result details https://arcpr…
ARC-AGI-1 semi-private; GPT-5.6 Sol (Medium); CoT; cost per task $0.22; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Solv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Sol (XHigh); CoT; cost per task $0.4; result details https://arcpriz…
ARC-AGI-1 semi-private; GPT-5.6 Sol (XHigh); CoT; cost per task $0.4; result details https://arcprize.org/results/openai-gpt-5-6-sol; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Terra (High); CoT; cost per task $0.19; result details https://arcpr…
ARC-AGI-1 semi-private; GPT-5.6 Terra (High); CoT; cost per task $0.19; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Terra (Low); CoT; cost per task $0.09; result details https://arcpri…
ARC-AGI-1 semi-private; GPT-5.6 Terra (Low); CoT; cost per task $0.09; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Terra (Max); CoT; cost per task $0.55; result details https://arcpri…
ARC-AGI-1 semi-private; GPT-5.6 Terra (Max); CoT; cost per task $0.55; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Terra (Medium); CoT; cost per task $0.13; result details https://arc…
ARC-AGI-1 semi-private; GPT-5.6 Terra (Medium); CoT; cost per task $0.13; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-5.6 Terrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-5.6 Terra (XHigh); CoT; cost per task $0.26; result details https://arcp…
ARC-AGI-1 semi-private; GPT-5.6 Terra (XHigh); CoT; cost per task $0.26; result details https://arcprize.org/results/openai-gpt-5-6-terra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.;…
ARC-AGI-1 semi-private; GPT-6 Astra (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.2837032; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; c…
ARC-AGI-1 semi-private; GPT-6 Astra (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.22650770000000006; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; c…
ARC-AGI-1 semi-private; GPT-6 Astra (Max); CoT; Evaluation conducted with 'max' reasoning effort.; cost per task $0.4328827000000001; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effo…
ARC-AGI-1 semi-private; GPT-6 Astra (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.2398342000000002; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.;…
ARC-AGI-1 semi-private; GPT-6 Astra (None); CoT; Evaluation conducted with 'none' reasoning effort.; cost per task $0.21427369999999996; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
GPT-6 Astrav1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort…
ARC-AGI-1 semi-private; GPT-6 Astra (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $0.3470552; result details https://arcprize.org/results/openai-gpt-6-astra; result date not supplied; source captured 2026-09-10
Grok 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.5 (High); CoT; cost per task $0.33; result details https://arcprize.o…
ARC-AGI-1 semi-private; Grok 4.5 (High); CoT; cost per task $0.33; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.5 (Low); CoT; cost per task $0.18; result details https://arcprize.or…
ARC-AGI-1 semi-private; Grok 4.5 (Low); CoT; cost per task $0.18; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.5v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.5 (Medium); CoT; cost per task $0.31; result details https://arcprize…
ARC-AGI-1 semi-private; Grok 4.5 (Medium); CoT; cost per task $0.31; result details https://arcprize.org/results/xai-grok-4-5; result date not supplied; source captured 2026-09-10
Grok 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.6 (High); CoT; Evaluation conducted with 'high' reasoning effort.; co…
ARC-AGI-1 semi-private; Grok 4.6 (High); CoT; Evaluation conducted with 'high' reasoning effort.; cost per task $0.3268656; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.6 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost…
ARC-AGI-1 semi-private; Grok 4.6 (Low); CoT; Evaluation conducted with 'low' reasoning effort.; cost per task $0.10347528000000006; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.6 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.…
ARC-AGI-1 semi-private; Grok 4.6 (Medium); CoT; Evaluation conducted with 'medium' reasoning effort.; cost per task $0.30111846000000003; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
Grok 4.6v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; …
ARC-AGI-1 semi-private; Grok 4.6 (XHigh); CoT; Evaluation conducted with 'xhigh' reasoning effort.; cost per task $0.348182; result details https://arcprize.org/results/xai-grok-4-6; result date not supplied; source captured 2026-09-10
o3v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3 (High); CoT; cost per task $0.5002; ; result date not supplied; source ca…
ARC-AGI-1 semi-private; o3 (High); CoT; cost per task $0.5002; ; result date not supplied; source captured 2026-09-10
o3v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3 (Low); CoT; cost per task $0.1764; ; result date not supplied; source cap…
ARC-AGI-1 semi-private; o3 (Low); CoT; cost per task $0.1764; ; result date not supplied; source captured 2026-09-10
o3v1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3 (Medium); CoT; cost per task $0.2882; ; result date not supplied; source …
ARC-AGI-1 semi-private; o3 (Medium); CoT; cost per task $0.2882; ; result date not supplied; source captured 2026-09-10
o3 Miniv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-mini (High); CoT; cost per task $0.3989; ; result date not supplied; sour…
ARC-AGI-1 semi-private; o3-mini (High); CoT; cost per task $0.3989; ; result date not supplied; source captured 2026-09-10
o3 Miniv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-mini (Low); CoT; cost per task $0.0519; ; result date not supplied; sourc…
ARC-AGI-1 semi-private; o3-mini (Low); CoT; cost per task $0.0519; ; result date not supplied; source captured 2026-09-10
o3 Miniv1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-mini (Medium); CoT; cost per task $0.1907; ; result date not supplied; so…
ARC-AGI-1 semi-private; o3-mini (Medium); CoT; cost per task $0.1907; ; result date not supplied; source captured 2026-09-10
o3 Prov1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-Pro (High); CoT + Synthesis; cost per task $4.16; ; result date not suppl…
ARC-AGI-1 semi-private; o3-Pro (High); CoT + Synthesis; cost per task $4.16; ; result date not supplied; source captured 2026-09-10
o3 Prov1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-Pro (Low); CoT + Synthesis; cost per task $1.6382; ; result date not supp…
ARC-AGI-1 semi-private; o3-Pro (Low); CoT + Synthesis; cost per task $1.6382; ; result date not supplied; source captured 2026-09-10
o3 Prov1_Semi_Private
ARC-AGI · ARC-AGI (reviewed publisher results)
Single-model result · Positions withheld
Measurement date not reported
Scope · ARC-AGI-1 semi-private; o3-Pro (Medium); CoT + Synthesis; cost per task $3.1766; ; result date not s…
ARC-AGI-1 semi-private; o3-Pro (Medium); CoT + Synthesis; cost per task $3.1766; ; result date not supplied; source captured 2026-09-10