Model results

Humanity's Last Exam

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.

Official source

← All evaluation groups

Humanity's Last Exam

Source: Humanity's Last Exam via Epoch AI
Version: Not reported
Measurement date: Not reported
Metric: % · Higher is better
Evaluation scope: effort unspecified · result date not supplied

Positions are withheld because at least one result lacks explicit independent provenance or a documented evaluation scope. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

GPT-6 Astra
openai/gpt-6-astra-20260903
54.8%
View modelEvidence aggregated
Gemini 3.8 Flash
google/gemini-3.8-flash-20260902
44.5%
View modelEvidence aggregated
GPT-5.4 Pro
openai/gpt-5.4-pro-20260305
44.3%
View modelEvidence aggregated
Claude Opus 4.7
anthropic/claude-opus-4-7
36.2%
View modelEvidence aggregated
GPT-5 Pro
openai/gpt-5-pro-2025-10-06
31.6%
View modelEvidence aggregated
Claude Haiku 5.5
anthropic/claude-haiku-5-5
30.1%
View modelEvidence aggregated
GPT-5.2
openai/gpt-5.2-20251211
27.8%
View modelEvidence aggregated
GPT-5
openai/gpt-5-2025-08-07
25.3%
View modelEvidence aggregated
GPT-5 Mini
openai/gpt-5-mini-2025-08-07
19.4%
View modelEvidence aggregated
Claude Opus 4.5
anthropic/claude-opus-4-5
14.2%
View modelEvidence aggregated
o1
openai/o1-2024-12-17
8%
View modelEvidence aggregated
Claude Opus 4.1
anthropic/claude-opus-4-1
7.9%
View modelEvidence aggregated
Claude Sonnet 4.5
anthropic/claude-sonnet-4-5
7.5%
View modelEvidence aggregated
Claude Opus 4
anthropic/claude-opus-4
6.7%
View modelEvidence aggregated
Claude Sonnet 4
anthropic/claude-sonnet-4
5.5%
View modelEvidence aggregated