Frontier Log

Benchmark · InactiveNo longer reported by the major labs since Jun 2026

IFEval

Created by Google and Yale University, 14 Nov 2023 · Higher is better · Website

020406080100Apr 2025Jul 2025Oct 2025Jan 2026Apr 2026Jul 202694.6%Qwen3.7-Plus94.3%Qwen3.7-Max85.2%Qwen3.5-Omni-Flash93.4%Qwen3-Max-Thinking88.9%Qwen3-Next83.4%Qwen3-235B-A22B020406080100Apr 2025Jul 2025Oct 2025Jan 2026Apr 2026Jul 202694.6%Qwen3.7-Plus85.2%Qwen3.5-Omni-Flash93.4%Qwen3-Max-Thinking88.9%Qwen3-Next83.4%Qwen3-235B-A22B

Results

ModelScoreSourceSource check
Qwen3.7-Plus1 Jun 2026 · Alibaba94.6%Qwen3.7-Plus: Multimodal Agent IntelligenceOur capture
Qwen3.7-Max20 May 2026 · Alibaba94.3%Qwen3.7: The Agent Frontier
Qwen3.5-Omni30 Mar 2026 · Alibaba89.7%Qwen3.5-Omni: Scaling Up, Toward Native Omni-Modal AGIQwen3.5-Omni-PlusOur capture
Qwen3.5-Omni-Flash30 Mar 2026 · Alibaba85.2%Qwen3.5-Omni: Scaling Up, Toward Native Omni-Modal AGIQwen3.5-Omni-FlashOur capture
Qwen3.5-397B-A17B16 Feb 2026 · Alibaba92.6%Qwen3.5: Towards Native Multimodal Agents
CorroboratedThe same number is printed in 1 other lab document: 1
Qwen3-Max-Thinking26 Jan 2026 · Alibaba93.4%Qwen3.5: Towards Native Multimodal Agentsreported later
CorroboratedThe same number is printed in 1 other lab document: 1
Qwen3-Next11 Sep 2025 · Alibaba88.9%Qwen/Qwen3-Next-80B-A3B-Thinking model card (launch-day revision)thinking modelOur capture
Qwen3-235B-A22B-Thinking-250725 Jul 2025 · Alibaba87.8%Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)Our capture
Mistral Medium 37 May 2025 · Mistral89.4%Medium is the new large.0-shotSource image
Qwen3-235B-A22B29 Apr 2025 · Alibaba83.4%Qwen/Qwen3-235B-A22B-Thinking-2507 model card (revision of 2025-07-25)thinking mode; reported laterOur capture

All benchmarks