supermarket

Language

Qwen3.5-397B-A17B

By Alibaba Group, in the Qwen family.

Released
16 Feb 2026
Status
Legacy
Context
262.1K
Max output
65.5K
Input per 1M tokens
$0.60
Output per 1M tokens
$3.60
Open weights
Yes
License
apache-2.0
Parameters
397B, 17B active
Knowledge cutoff
—
Reasoning
Yes
Modalities
text, image, video → text

Category scores

Out of 100, from independent and third-party results only, by method category-2026-09-29.1.

Category scores
CategoryScoreRankCoverage
Knowledge & reasoning26.5652 of 4
Human preference53.0Not ranked2 of 5

Benchmark results

Every result we hold for this model, with who ran it and where it was published.

Knowledge & reasoning

Knowledge & reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Chess PuzzlesPuzzles solved13.0%Independentnone7 Aug 2026Epoch AI
Chess PuzzlesPuzzles solved12.0%Independent—7 Aug 2026Epoch AI
DTBench (Decision Theory Benchmark)Accuracy (capability questions)87.5%Third party——https://conceptualreasoning.ai/dtbench via Epoch AI
Humanity's Last Exam (HLE)Accuracy48.3%Self-reported——Model Card via Hugging Face Community Evals
LMCA (Language Model Conceptual Argumentation)Correlation with expert ratings (Pearson r)0.3789Third party——https://conceptualreasoning.ai/lmca via Epoch AI
MMLU-ProAccuracy87.8%Self-reported——Model Card via Hugging Face Community Evals
Mystery Game PuzzlesPuzzles solved18.0%Independentnone27 Aug 2026Epoch AI

Scientific reasoning

Scientific reasoning results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
GPQA DiamondAccuracy86.4%Independentnone7 Aug 2026Epoch AI
GPQA DiamondAccuracy85.9%Independent—7 Aug 2026Epoch AI
GPQA DiamondAccuracy88.4%Self-reported——Model Card via Hugging Face Community Evals

Mathematics

Mathematics results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
FrontierMath Tiers 1-3Accuracy31.2%Independentnone30 Aug 2026Epoch AI
FrontierMath Tiers 1-3Accuracy29.5%Independent—28 Aug 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy88.9%Independent—7 Aug 2026Epoch AI
OTIS Mock AIME 2024-2025Accuracy82.2%Independentnone7 Aug 2026Epoch AI

Professional knowledge work

Professional knowledge work results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
APEX-AgentsPass@1 (all rubric criteria met)24.9%Source not stated——Epoch AI via Epoch AI

Human preference

Human preference results
BenchmarkMetricValueProvenanceSettingEvaluatedSource
Code Arena: WebDev (arena.ai)Arena score (Elo scale)1400Independent—25 Sep 2026Arena
Code Arena: WebDev (arena.ai)Votes21305Independent—25 Sep 2026Arena
Text Arena (arena.ai)Arena score (Elo scale)1442Independent—25 Sep 2026Arena
Text Arena (arena.ai)Votes86319Independent—25 Sep 2026Arena
Vision Arena (arena.ai)Arena score (Elo scale)1245Independent—28 Sep 2026Arena
Vision Arena (arena.ai)Votes31178Independent—28 Sep 2026Arena

Other Qwen versions

Sources