Qwen
Qwen
Language model · China ·Apr 2025

Qwen3-32B

Language Open (Restricted)
32.8B
Parameters
60
Benchmarks

About Qwen3-32B

Qwen3-32B is an AI model developed by Qwen, in the language category, released in 2025, made available as an open-weights model with 32.8B params.

On this page you'll find Qwen3-32B's full specifications, including 60 benchmark results. Review provider pricing and benchmark scores below, or compare Qwen3-32B head-to-head with other language models.

Links

Run Qwen3-32B locally — what to buy

Share this build ↗

Sized to what Qwen3-32B actually needs (~8k context at Q4), not to the biggest GPU: Good = cheapest that runs it, Better = best value, Best = most headroom. Speed is a hardware estimate; anything past ~120 tok/s is shown as instant.

Software support

✓ measured · · compatible · — not supported. Informational only — speed is hardware-based.

Community

Would you run Qwen3-32B again?
No votes yet — be the first.
Did it run? — community reports by hardware

No reports yet. Own this model on some hardware? Be the first to confirm it runs.

Discussion

No comments yet. Start the discussion.

Qwen3-32B benchmark scores
Benchmark Category Score Variant
Codeforces Elo code 1,977 thinking-rating
Codeforces Elo code 1,353 non-thinking-rating
LiveCodeBench v5 code 65.7% thinking-pass@1
LiveCodeBench v5 code 65.7% reasoning
LiveCodeBench v5 code 53.24% thinking-cited-ntele-readme
LiveCodeBench v5 code 31.3% non-thinking-pass@1
LiveCodeBench v6 code 60.1% reasoning
AlignBench v1.1 general 8.72 thinking-v1-1
AlignBench v1.1 general 8.58 non-thinking-v1-1
ArenaHard general 93.8% thinking
ArenaHard general 92.8% non-thinking
C-Eval general 87.3% thinking
C-Eval general 83.3% non-thinking
Creative Writing v3 general 81 thinking
Creative Writing v3 general 78.3 non-thinking
IFEval general 85% thinking-strict-prompt
IFEval general 85% reasoning
IFEval general 83.2% non-thinking-strict-prompt
LiveBench general 74.9 thinking-2024-11-25
LiveBench general 59.8 non-thinking-2024-11-25
MMLU-Pro general 80% reasoning
MMLU-Redux general 90.9% thinking
MMLU-Redux general 90.9% reasoning
MMLU-Redux general 85.7% non-thinking
MMMLU general 80.6% thinking-14lang
MMMLU general 76.5% non-thinking-14lang
Multi-IF general 73 thinking
Multi-IF general 70.7 non-thinking
WritingBench general 7.9 thinking
WritingBench general 7.54 non-thinking
INCLUDE knowledge 73.7 thinking
INCLUDE knowledge 70.9 non-thinking
AIME 2024 math 81.4% thinking
AIME 2024 math 78.75% thinking-cited-ntele-readme
AIME 2024 math 31% non-thinking
AIME 2025 math 73.33% thinking-cited-ntele-readme
AIME 2025 math 72.9% thinking
AIME 2025 math 72.9% reasoning
AIME 2025 math 20.2% non-thinking
HMMT February 2025 math 50.4% reasoning
MATH-500 math 97.2% thinking
MATH-500 math 95% thinking-cited-ntele-readme
MATH-500 math 88.6% non-thinking
MT-AIME 2024 math 75 thinking-mt-2024
MT-AIME 2024 math 24.1 non-thinking-mt-2024
PolyMath-en math 47.4 thinking
PolyMath-en math 22.5 non-thinking
AutoLogi reasoning 87.3 thinking
AutoLogi reasoning 78.5 non-thinking
GPQA Diamond reasoning 69.7% thinking-cited-ntele-readme
GPQA Diamond reasoning 68.4% thinking
GPQA Diamond reasoning 68.4% reasoning
GPQA Diamond reasoning 54.6% non-thinking
MLogiQA reasoning 76.3 thinking
MLogiQA reasoning 62.9 non-thinking
ZebraLogic reasoning 88.8% thinking
ZebraLogic reasoning 29.2% non-thinking
BFCL v3 tool-use 70.3% reasoning
BFCL v3 tool-use 70.3% thinking
BFCL v3 tool-use 63% non-thinking

More models like Qwen3-32B

Frequently asked questions

What is Qwen3-32B?

Qwen3-32B is an AI model developed by Qwen, in the language category, released in 2025. It is tracked on GenAIList with its specifications, benchmark scores and provider pricing.

Who created Qwen3-32B?

Qwen3-32B was developed by Qwen and released in 2025.

Is Qwen3-32B open source or proprietary?

Qwen3-32B ships as an open-weights model: you can download and self-host the weights, though the licence may place some restrictions on use.

How much does Qwen3-32B cost?

Pricing for Qwen3-32B depends on the provider. See the providers table on this page for the latest API rates.

How does Qwen3-32B perform on benchmarks?

Qwen3-32B is benchmarked across 60 evaluations on GenAIList, including Codeforces Elo (1,977). See the full benchmark table below and compare it with other models.

No videos yet. Know a good one? Add it below.

Know a great video about Qwen3-32B?