Model results

τ²-bench

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

τ²-bench

Source: Artificial Analysis
Version: Not reported
Measurement date: Not reported
Metric: % · Higher is better
Evaluation scope: Tool-agent benchmark across supported domains; observed 2026-10-19; result date not supplied

Exact score ties share a position. Positions do not establish statistical significance. Missing metadata is shown as unreported; matching imported metadata is not independent verification of the protocol.

#1Claude Fable 5
anthropic/claude-fable-5
98.5%
View modelEvidence independent
#2MiMo-V2.5-Pro
xiaomi/mimo-v2.5-pro-20260422
94.2%
View modelEvidence independent
#2Mistral Medium 3
mistralai/mistral-medium-3
94.2%
View modelEvidence independent
#2Mistral Medium 3.1
mistralai/mistral-medium-3.1
94.2%
View modelEvidence independent
#2Mistral Medium 3.5
mistralai/mistral-medium-3.5-20260430
94.2%
View modelEvidence independent
#6MiMo-V2.5
xiaomi/mimo-v2.5-20260422
90.6%
View modelEvidence independent
#7MiniMax M3
minimax/minimax-m3-20260531
88.9%
View modelEvidence independent
#8GPT-5.6 Terra
openai/gpt-5.6-terra-20260709
86.3%
View modelEvidence independent
#9GPT-5.6 Sol
openai/gpt-5.6-sol-20260709
85.1%
View modelEvidence independent
#10MiMo-V2-Flash
xiaomi/mimo-v2-flash
83.9%
View modelEvidence independent
#11Nemotron 3 Ultra
nvidia/nemotron-3-ultra-550b-a55b-20260604
83.3%
View modelEvidence independent
#12Command A
cohere/command-a-03-2025
80.7%
View modelEvidence independent
#12Command A+
cohere/command-a-plus-05-2026
80.7%
View modelEvidence independent
#14gpt-oss-120b
openai/gpt-oss-120b
65.8%
View modelEvidence independent
#15Phi 4
microsoft/phi-4
0%
View modelEvidence independent