evals.report
BenchmarksLabsCompareRun guidesIn the wild
Alibaba / QwenQwen

Qwen3 235B A22B Instruct 2507

Alibaba / Qwen · Qwen. Released Jul 21, 2025.

Qwen3 235B A22B Instruct 2507 is a model from Alibaba / Qwen in the Qwen family, released Jul 21, 2025. evals.report tracks 21 reported Qwen3 235B A22B Instruct 2507 benchmark scores across Berkeley Function Calling Leaderboard, LiveCodeBench Pro, SWE-bench Pro, SimpleQA Verified, GPQA Diamond, FrontierMath, AIME (OTIS Mock), LMArena, and 13 more — each shown with its benchmark, metric, source status, and date, and never combined into a single ranking.

Open21 results

Benchmark results 21

Compare this model
BenchmarkCategoryScoreMetricStatusDate
Berkeley Function Calling LeaderboardTool use52.15%accuracyOfficialJul 21, 2025Details
LiveCodeBench ProCoding1673Codeforces EloOfficialJul 21, 2025Details
SWE-bench ProCoding21.41%% resolvedOfficialJul 21, 2025Details
SimpleQA VerifiedOther50.1%accuracyOfficialJul 21, 2025Details
GPQA DiamondReasoning80.1%accuracyOfficialJul 21, 2025Details
FrontierMathReasoning8.48%accuracyOfficialJul 21, 2025Details
AIME (OTIS Mock)Reasoning86.7%accuracyOfficialJul 21, 2025Details
LMArenaChat preference1419source-defined ratingOfficialJul 21, 2025Details
WeirdMLCoding38.7%average accuracyOfficialJul 21, 2025Details
MCP AtlasTool use12.0%pass rateOfficialJul 21, 2025Details
Epoch Capabilities IndexReasoning139.1IndexOfficialJul 21, 2025Details
Aider PolyglotCoding57.3%% correctUnverifiedJul 21, 2025Details
MMLU-ProReasoning82.8%accuracyVerifiedJul 21, 2025Details
GDPvalAgents778EloOfficialJul 21, 2025Details
LiveCodeBenchCoding52.4%Pass@1UnverifiedJul 21, 2025Details
SciCodeCoding36.0%accuracyUnverifiedJul 21, 2025Details
LongBench v2Reasoning58.3%accuracyOfficialJul 21, 2025Details
Design ArenaChat preference1093EloVerifiedJul 21, 2025Details
MCP-UniverseTool use18.18%Overall Success RateVerifiedJul 21, 2025Details
SuperGPQAReasoning62.6%accuracyUnverifiedJul 21, 2025Details
IMO-BenchReasoning53.8%accuracyVerifiedJul 21, 2025Details