Qwen
Qwen
Language model · China ·Apr 2025

Qwen3-8B

Language Open (Restricted)
8.2B
Parameters
46
Benchmarks

About Qwen3-8B

Qwen3-8B is an AI model developed by Qwen, in the language category, released in 2025, made available as an open-weights model with 8.2B params.

On this page you'll find Qwen3-8B's full specifications, including 46 benchmark results. Review provider pricing and benchmark scores below, or compare Qwen3-8B head-to-head with other language models.

Links

Run Qwen3-8B locally — what to buy

Share this build ↗

Sized to what Qwen3-8B 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 Qwen3-8B 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.

Qwen3-8B benchmark scores
Benchmark Category Score Variant
Codeforces Elo code 1,785 thinking-rating
Codeforces Elo code 1,110 non-thinking-rating
LiveCodeBench v5 code 54.5% thinking-pass@1
LiveCodeBench v5 code 22.8% non-thinking-pass@1
AlignBench v1.1 general 8.46 thinking-v1-1
AlignBench v1.1 general 8.38 non-thinking-v1-1
ArenaHard general 85.8% thinking
ArenaHard general 79.6% non-thinking
C-Eval general 83.4% thinking
C-Eval general 77.9% non-thinking
Creative Writing v3 general 75 thinking
Creative Writing v3 general 64.5 non-thinking
IFEval general 85% thinking-strict-prompt
IFEval general 83% non-thinking-strict-prompt
LiveBench general 67.1 thinking-2024-11-25
LiveBench general 53.5 non-thinking-2024-11-25
MMLU-Redux general 87.5% thinking
MMLU-Redux general 79.5% non-thinking
MMMLU general 74.4% thinking-14lang
MMMLU general 66.9% non-thinking-14lang
Multi-IF general 71.2 thinking
Multi-IF general 69.2 non-thinking
WritingBench general 7.59 thinking
WritingBench general 7.15 non-thinking
INCLUDE knowledge 67.8 thinking
INCLUDE knowledge 62.5 non-thinking
AIME 2024 math 76% thinking
AIME 2024 math 29.1% non-thinking
AIME 2025 math 67.3% thinking
AIME 2025 math 20.9% non-thinking
MATH-500 math 97.4% thinking
MATH-500 math 87.4% non-thinking
MT-AIME 2024 math 65.4 thinking-mt-2024
MT-AIME 2024 math 16.6 non-thinking-mt-2024
PolyMath-en math 42.7 thinking
PolyMath-en math 18.8 non-thinking
AutoLogi reasoning 89.1 thinking
AutoLogi reasoning 76.5 non-thinking
GPQA Diamond reasoning 62% thinking
GPQA Diamond reasoning 39.3% non-thinking
MLogiQA reasoning 69 thinking
MLogiQA reasoning 51.4 non-thinking
ZebraLogic reasoning 84.8% thinking
ZebraLogic reasoning 26.7% non-thinking
BFCL v3 tool-use 68.1% thinking
BFCL v3 tool-use 60.2% non-thinking

More models like Qwen3-8B

Frequently asked questions

What is Qwen3-8B?

Qwen3-8B 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 Qwen3-8B?

Qwen3-8B was developed by Qwen and released in 2025.

Is Qwen3-8B open source or proprietary?

Qwen3-8B 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 Qwen3-8B cost?

Pricing for Qwen3-8B depends on the provider. See the providers table on this page for the latest API rates.

How does Qwen3-8B perform on benchmarks?

Qwen3-8B is benchmarked across 46 evaluations on GenAIList, including Codeforces Elo (1,785). 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 Qwen3-8B?