Toolathlon-Verified

A verified tool-use benchmark variant for completing multi-step workflows with external tools.

In index
Agentic
Best in agentic

Models23

Top result80.6%Claude Opus 5

Top-3 spread2.8 pts

Year—

Result and price

16 of 23 results

Results

23 results

#ModelResult

01Claude Opus 580.6%

02GLM-5.3-Flash78.4%

03Claude Opus 5.577.8%

03Claude Sonnet 5.577.8%

03Claude Fable 5.177.8%

06MiMo-V2.6-Pro76.9%

07ERNIE 5.5 Flash Preview75.0%

08Step 5 Preview74.1%

08Hy4 preview74.1%

08DeepSeek V4 Pro 081374.1%

11MiMo-V2.6-Flash73.6%

12Qwen3.8-Flash-Next73.5%

13Kimi K373.2%

14GLM-5.373.0%

15Muse Spark Open72.8%

16Qwen3.8 Max72.5%

17Ornith-1.5-397B71.2%

18DeepSeek V4 Flash 073170.3%

19dots3-note Preview55.6%

20Inkling-Small54.4%

21Laguna S 2.149.7%

22Ornith-1.5-35B-A3B48.7%

23Ornith-1.5-9B41.2%