Skip to content
AI Atlas
BenchmarkActive

τ²-bench

dual-control tool-agent-user interaction (telecom, retail, airline)

quality51

Updated 10 h ago · first seen 11 Sept 2026

Metric
— · % ↑
Current results
0
Models
0
Current leader
—

No results recorded for this benchmark yet — its sources are being connected. The definition, aliases and variants are kept so links resolve; nothing is fabricated.

Score history · Qwen3 32B 1 row

Not enough history to chart — a single observation (29.82% on 11 Sept 2026). Rows under different configurations count separately; the list below shows each one.

  • 29.82%aa_slug=qwen3-32b-instruct-reasoning · evaluator=Artificial Analysis · index_version=4.311 Sept 2026

Back to the leaderboard

Leaderboard 440 models

Select models with +, then Compare.

Leaderboard
#ModelScoreTrustConfigurationvs leaderEvaluatedSourceActions
299hermes-4-llama-3-1-405bOpen weightsNous Research26.6%Independentgroup defaultsComparable0.00 ptobs. 11 Sept 2026artificialanalysis.aiT2History
299intellect-3Open weightsPrime Intellect26.6%Independentgroup defaultsComparable0.00 ptobs. 11 Sept 2026artificialanalysis.aiT2History
299magistral-smallOpen weightsMistral AI26.6%Independentgroup defaultsComparable0.00 ptobs. 11 Sept 2026artificialanalysis.aiT2History
299qwen3-4b-2507-instructOpen weightsAlibaba Group26.6%Independentgroup defaultsComparable0.00 ptobs. 11 Sept 2026artificialanalysis.aiT2History
305ring-1tOpen weightsinclusionAI26.3%Independentgroup defaultsComparable-0.29 ptobs. 11 Sept 2026artificialanalysis.aiT2History
306gemma-4-e4b-non-reasoningOpen weightsGoogle26.0%Independentgroup defaultsComparable-0.59 ptobs. 11 Sept 2026artificialanalysis.aiT2History
306qwen3-1.7b-instruct-reasoningOpen weightsAlibaba Group26.0%Independentgroup defaultsComparable-0.59 ptobs. 11 Sept 2026artificialanalysis.aiT2History
306qwen3-30b-a3b-instruct-reasoningOpen weightsAlibaba Group26.0%Independentgroup defaultsComparable-0.59 ptobs. 11 Sept 2026artificialanalysis.aiT2History
309gpt-5-nano-minimalClosedOpenAI25.7%Independentgroup defaultsComparable-0.88 ptobs. 11 Sept 2026artificialanalysis.aiT2History
310Nemotron 3 Nano 30B A3BOpen weightsNVIDIA25.4%Independentgroup defaultsComparable-1.17 ptobs. 11 Sept 2026artificialanalysis.aiT2History
310k2-think-v2Open weightsMBZUAI Institute of Foundation Models25.4%Independentgroup defaultsComparable-1.17 ptobs. 11 Sept 2026artificialanalysis.aiT2History
310qwen3-4b-2507-instruct-reasoningOpen weightsAlibaba Group25.4%Independentgroup defaultsComparable-1.17 ptobs. 11 Sept 2026artificialanalysis.aiT2History
313Mistral Small 3.1Open weightsMistral AI25.1%Independentgroup defaultsComparable-1.46 ptobs. 11 Sept 2026artificialanalysis.aiT2History
313gpt-4oClosedOpenAI25.1%Independentgroup defaultsComparable-1.46 ptobs. 11 Sept 2026artificialanalysis.aiT2History
313llama-nemotron-super-49b-v1-5Open weightsNVIDIA25.1%Independentgroup defaultsComparable-1.46 ptobs. 11 Sept 2026artificialanalysis.aiT2History
316Devstral 2Open weightsMistral AI24.9%Independentgroup defaultsComparable-1.76 ptobs. 11 Sept 2026artificialanalysis.aiT2History
316Ministral 3 3BOpen weightsMistral AI24.9%Independentgroup defaultsComparable-1.76 ptobs. 11 Sept 2026artificialanalysis.aiT2History
316k2-v2-mediumOpen weightsMBZUAI Institute of Foundation Models24.9%Independentgroup defaultsComparable-1.76 ptobs. 11 Sept 2026artificialanalysis.aiT2History
316qwen3-8b-instructOpen weightsAlibaba Group24.9%Independentgroup defaultsComparable-1.76 ptobs. 11 Sept 2026artificialanalysis.aiT2History
320Claude Haiku 3.5ClosedAnthropic · Claude24.6%Independentgroup defaultsComparable-2.05 ptobs. 11 Sept 2026artificialanalysis.aiT2History
320Mistral Large 3Open weightsMistral AI24.6%Independentgroup defaultsComparable-2.05 ptobs. 11 Sept 2026artificialanalysis.aiT2History
322Mistral Medium 3ClosedMistral AI24.3%Independentgroup defaultsComparable-2.34 ptobs. 11 Sept 2026artificialanalysis.aiT2History
323qwen3-235b-a22b-instruct-reasoningOpen weightsAlibaba Group24.0%Independentgroup defaultsComparable-2.63 ptobs. 11 Sept 2026artificialanalysis.aiT2History
324Devstral Small 2Open weightsMistral AI23.4%Independentgroup defaultsComparable-3.22 ptobs. 11 Sept 2026artificialanalysis.aiT2History
324Qwen3-VL-4B-InstructOpen weightsQwen23.4%Independentgroup defaultsComparable-3.22 ptobs. 11 Sept 2026artificialanalysis.aiT2History
324gpt-5-4-mini-non-reasoningClosedOpenAI23.4%Independentgroup defaultsComparable-3.22 ptobs. 11 Sept 2026artificialanalysis.aiT2History
324nvidia-nemotron-nano-9b-v2Open weightsNVIDIA23.4%Independentgroup defaultsComparable-3.22 ptobs. 11 Sept 2026artificialanalysis.aiT2History
328llama-3-1-nemotron-instruct-70bOpen weightsNVIDIA23.1%Independentgroup defaultsComparable-3.51 ptobs. 11 Sept 2026artificialanalysis.aiT2History
328magistral-mediumClosedMistral AI23.1%Independentgroup defaultsComparable-3.51 ptobs. 11 Sept 2026artificialanalysis.aiT2History
330deepseek-v3Open weightsDeepSeek22.8%Independentgroup defaultsComparable-3.80 ptobs. 11 Sept 2026artificialanalysis.aiT2History
330granite-4-0-nano-1bOpen weightsIBM22.8%Independentgroup defaultsComparable-3.80 ptobs. 11 Sept 2026artificialanalysis.aiT2History
332GLM 4.5VOpen weightsZ.ai (Zhipu AI)22.5%Independentgroup defaultsComparable-4.10 ptobs. 11 Sept 2026artificialanalysis.aiT2History
332Hermes-4-70BOpen weightsNous Research22.5%Independentgroup defaultsComparable-4.10 ptobs. 11 Sept 2026artificialanalysis.aiT2History
332qwen3-vl-8b-reasoningOpen weightsAlibaba Group22.5%Independentgroup defaultsComparable-4.10 ptobs. 11 Sept 2026artificialanalysis.aiT2History
335Hermes 4 405BOpen weightsNous Research22.2%Independentgroup defaultsComparable-4.39 ptobs. 11 Sept 2026artificialanalysis.aiT2History
335gemma-4-e2b-non-reasoningOpen weightsGoogle22.2%Independentgroup defaultsComparable-4.39 ptobs. 11 Sept 2026artificialanalysis.aiT2History
335qwen3-30b-a3b-instructOpen weightsAlibaba Group22.2%Independentgroup defaultsComparable-4.39 ptobs. 11 Sept 2026artificialanalysis.aiT2History
338NVIDIA-Nemotron-Nano-9B-v2Open weightsNVIDIA21.9%Independentgroup defaultsComparable-4.68 ptobs. 11 Sept 2026artificialanalysis.aiT2History
338R1 Distill Llama 70BOpen weightsDeepSeek21.9%Independentgroup defaultsComparable-4.68 ptobs. 11 Sept 2026artificialanalysis.aiT2History
340Qwen3 Next 80B A3B InstructOpen weightsQwen21.6%Independentgroup defaultsComparable-4.97 ptobs. 11 Sept 2026artificialanalysis.aiT2History
340hermes-4-llama-3-1-70bOpen weightsNous Research21.6%Independentgroup defaultsComparable-4.97 ptobs. 11 Sept 2026artificialanalysis.aiT2History
340nanbeige4-1-3bOpen weightsNanbeige21.6%Independentgroup defaultsComparable-4.97 ptobs. 11 Sept 2026artificialanalysis.aiT2History
340qwen3-1.7b-instructOpen weightsAlibaba Group21.6%Independentgroup defaultsComparable-4.97 ptobs. 11 Sept 2026artificialanalysis.aiT2History
344nvidia-nemotron-nano-12b-v2-vl-reasoningOpen weightsNVIDIA21.4%Independentgroup defaultsComparable-5.26 ptobs. 11 Sept 2026artificialanalysis.aiT2History
344olmo-3-1-32b-instructOpen weightsAllen Institute for AI21.4%Independentgroup defaultsComparable-5.26 ptobs. 11 Sept 2026artificialanalysis.aiT2History
344qwen3-omni-30b-a3b-reasoningOpen weightsAlibaba Group21.4%Independentgroup defaultsComparable-5.26 ptobs. 11 Sept 2026artificialanalysis.aiT2History
347Claude 3 HaikuClosedAnthropic · Claude21.1%Independentgroup defaultsComparable-5.56 ptobs. 11 Sept 2026artificialanalysis.aiT2History
347Llama-3.2-3BRestricted weightsMeta AI21.1%Independentgroup defaultsComparable-5.56 ptobs. 11 Sept 2026artificialanalysis.aiT2History
347Qwen3-0.6BOpen weightsQwen21.1%Independentgroup defaultsComparable-5.56 ptobs. 11 Sept 2026artificialanalysis.aiT2History
350gemma-4-E2BOpen weightsGoogle20.8%Independentgroup defaultsComparable-5.85 ptobs. 11 Sept 2026artificialanalysis.aiT2History
350gemma-4-E4BOpen weightsGoogle20.8%Independentgroup defaultsComparable-5.85 ptobs. 11 Sept 2026artificialanalysis.aiT2History
350k2-v2-lowOpen weightsMBZUAI Institute of Foundation Models20.8%Independentgroup defaultsComparable-5.85 ptobs. 11 Sept 2026artificialanalysis.aiT2History
350ling-flash-2-0Open weightsinclusionAI20.8%Independentgroup defaultsComparable-5.85 ptobs. 11 Sept 2026artificialanalysis.aiT2History
354exaone-4-0-1-2bOpen weightsLG AI Research20.5%Independentgroup defaultsComparable-6.14 ptobs. 11 Sept 2026artificialanalysis.aiT2History
355solar-miniOpen weightsUpstage20.2%Independentgroup defaultsComparable-6.43 ptobs. 11 Sept 2026artificialanalysis.aiT2History
356devstral-mediumClosedMistral AI19.9%Independentgroup defaultsComparable-6.73 ptobs. 11 Sept 2026artificialanalysis.aiT2History
356qwen3-vl-30b-a3b-reasoningOpen weightsAlibaba Group19.9%Independentgroup defaultsComparable-6.73 ptobs. 11 Sept 2026artificialanalysis.aiT2History
358Mistral Small 3Open weightsMistral AI19.6%Independentgroup defaultsComparable-7.02 ptobs. 11 Sept 2026artificialanalysis.aiT2History
358glm-4-5vOpen weightsZ.ai (Zhipu AI)19.6%Independentgroup defaultsComparable-7.02 ptobs. 11 Sept 2026artificialanalysis.aiT2History
358granite-4-0-h-nano-1bOpen weightsIBM19.6%Independentgroup defaultsComparable-7.02 ptobs. 11 Sept 2026artificialanalysis.aiT2History
358granite-4.1-3bOpen weightsIBM19.6%Independentgroup defaultsComparable-7.02 ptobs. 11 Sept 2026artificialanalysis.aiT2History
358lfm2-5-1-2b-thinkingOpen weightsLiquid AI19.6%Independentgroup defaultsComparable-7.02 ptobs. 11 Sept 2026artificialanalysis.aiT2History
363nvidia-nemotron-nano-12b-v2-vlOpen weightsNVIDIA19.3%Independentgroup defaultsComparable-7.31 ptobs. 11 Sept 2026artificialanalysis.aiT2History
364Llama-3.1-405BRestricted weightsMeta AI19.0%Independentgroup defaultsComparable-7.60 ptobs. 11 Sept 2026artificialanalysis.aiT2History
364Qwen3 VL 30B A3B InstructOpen weightsQwen19.0%Independentgroup defaultsComparable-7.60 ptobs. 11 Sept 2026artificialanalysis.aiT2History
364Qwen3-4BOpen weightsQwen19.0%Independentgroup defaultsComparable-7.60 ptobs. 11 Sept 2026artificialanalysis.aiT2History
364gemini-2-5-flash-liteClosedGoogle19.0%Independentgroup defaultsComparable-7.60 ptobs. 11 Sept 2026artificialanalysis.aiT2History
368Gemini 2.5 Flash-LiteClosedGoogle · Gemini18.4%Independentgroup defaultsComparable-8.19 ptobs. 11 Sept 2026artificialanalysis.aiT2History
368mistral-small-4-non-reasoningOpen weightsMistral AI18.4%Independentgroup defaultsComparable-8.19 ptobs. 11 Sept 2026artificialanalysis.aiT2History
370Llama 4 MaverickOpen weightsMeta AI17.8%Independentgroup defaultsComparable-8.77 ptobs. 11 Sept 2026artificialanalysis.aiT2History
371nova-liteClosedAmazon Web Services17.5%Independentgroup defaultsComparable-9.07 ptobs. 11 Sept 2026artificialanalysis.aiT2History
372exaone-4-0-32b-reasoningOpen weightsLG AI Research17.3%Independentgroup defaultsComparable-9.36 ptobs. 11 Sept 2026artificialanalysis.aiT2History
372gpt-4.1-nanoClosedOpenAI17.3%Independentgroup defaultsComparable-9.36 ptobs. 11 Sept 2026artificialanalysis.aiT2History
372granite-4-0-h-smallOpen weightsIBM17.3%Independentgroup defaultsComparable-9.36 ptobs. 11 Sept 2026artificialanalysis.aiT2History
375Llama 3.1 8BRestricted weightsMeta AI16.4%Independentgroup defaultsComparable-10.2 ptobs. 11 Sept 2026artificialanalysis.aiT2History
375exaone-4-0-1-2b-reasoningOpen weightsLG AI Research16.4%Independentgroup defaultsComparable-10.2 ptobs. 11 Sept 2026artificialanalysis.aiT2History
375qwen3-omni-30b-a3b-instructOpen weightsAlibaba Group16.4%Independentgroup defaultsComparable-10.2 ptobs. 11 Sept 2026artificialanalysis.aiT2History
378LFM2.5-8B-A1BOpen weightsLiquid AI16.1%Independentgroup defaultsComparable-10.5 ptobs. 11 Sept 2026artificialanalysis.aiT2History
378step-3-vl-10bOpen weightsStepFun16.1%Independentgroup defaultsComparable-10.5 ptobs. 11 Sept 2026artificialanalysis.aiT2History
380jamba-reasoning-3bOpen weightsAI21 Labs15.8%Independentgroup defaultsComparable-10.8 ptobs. 11 Sept 2026artificialanalysis.aiT2History
381Llama 4 ScoutRestricted weightsMeta AI15.5%Independentgroup defaultsComparable-11.1 ptobs. 11 Sept 2026artificialanalysis.aiT2History
381qwen3-vl-4b-reasoningOpen weightsAlibaba Group15.5%Independentgroup defaultsComparable-11.1 ptobs. 11 Sept 2026artificialanalysis.aiT2History
383Command AOpen weightsCohere15.2%Independentgroup defaultsComparable-11.4 ptobs. 11 Sept 2026artificialanalysis.aiT2History
383Llama-3.1-70BOpen weightsMeta AI15.2%Independentgroup defaultsComparable-11.4 ptobs. 11 Sept 2026artificialanalysis.aiT2History
385Gemini 2.5 FlashClosedGoogle · Gemini14.9%Independentgroup defaultsComparable-11.7 ptobs. 11 Sept 2026artificialanalysis.aiT2History
386granite-4-0-h-350mOpen weightsIBM14.6%Independentgroup defaultsComparable-12.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
386llama-3-2-instruct-11b-visionOpen weightsMeta AI14.6%Independentgroup defaultsComparable-12.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
386qwen3-0.6b-instructOpen weightsAlibaba Group14.6%Independentgroup defaultsComparable-12.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
389nova-microClosedAmazon Web Services14.0%Independentgroup defaultsComparable-12.6 ptobs. 11 Sept 2026artificialanalysis.aiT2History
389nova-proClosedAmazon Web Services14.0%Independentgroup defaultsComparable-12.6 ptobs. 11 Sept 2026artificialanalysis.aiT2History
391jamba-1-7-largeOpen weightsAI21 Labs13.4%Independentgroup defaultsComparable-13.2 ptobs. 11 Sept 2026artificialanalysis.aiT2History
391lfm2-2-6bOpen weightsLiquid AI13.4%Independentgroup defaultsComparable-13.2 ptobs. 11 Sept 2026artificialanalysis.aiT2History
393granite-4-0-350mOpen weightsIBM13.2%Independentgroup defaultsComparable-13.4 ptobs. 11 Sept 2026artificialanalysis.aiT2History
393ling-mini-2-0Open weightsinclusionAI13.2%Independentgroup defaultsComparable-13.4 ptobs. 11 Sept 2026artificialanalysis.aiT2History
395apertus-70b-instructOpen weightsSwiss AI Initiative12.9%Independentgroup defaultsComparable-13.7 ptobs. 11 Sept 2026artificialanalysis.aiT2History
396Granite 4.0 MicroOpen weightsIBM12.6%Independentgroup defaultsComparable-14.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
396LFM2-1.2BOpen weightsLiquid AI12.6%Independentgroup defaultsComparable-14.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
396Olmo-3-7B-InstructOpen weightsAllen Institute for AI12.6%Independentgroup defaultsComparable-14.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
396jamba-1-7-miniOpen weightsAI21 Labs12.6%Independentgroup defaultsComparable-14.0 ptobs. 11 Sept 2026artificialanalysis.aiT2History
400llama-3-1-nemotron-nano-4b-reasoningOpen weightsNVIDIA11.7%Independentgroup defaultsComparable-14.9 ptobs. 11 Sept 2026artificialanalysis.aiT2History

One row per canonical model — its best current row inside this comparability group (effort variants are folded into the model). Bars are relative to the page's best score. “vs leader” reads comparability: partially comparable = same task, conditions differ (reasoning effort, temperature, judge). Rules →