models · qwen

Qwen3-4B

Compact 4B Qwen3 with hybrid thinking. Outperforms much larger models on math benchmarks.

What this registry has checked

Quarantine. We hold the publisher’s own published hash for these weights and have recorded it against this entry. We have not downloaded and hashed the weights ourselves, so the hash is the publisher’s claim, not our measurement.

Weight file SHA-256	6912e3a1fafaa368b0b53a0c13fee4803fd0c8017049157d15a89cbe996af020
Method	hf-published-sha256

Re-check this yourself → Source → Full record →

What is recorded

Publisher	qwen
Origin	China
Licence	Apache-2.0
Parameters	4,022,468,096 (4.02B)
Architecture	qwen3
Layers	36
Hidden size	2560
Attention	GQA (32 heads, 8 KV)
Native precision	BF16
Pulls recorded here	67,020
Downloads reported upstream	6,204,889

Reported scores

MMLU-Pro	58.6
GPQA Diamond	39.8
AIME 2025	58.3
MATH-500	84.3
LiveCodeBench	23.3
Humanity's Last Exam	3.3

Reported by a third party and recorded here, not re-run by us. Source: artificialanalysis.ai/api/v2.

Elsewhere in the registry

All models → Browse with filters → What will run on my hardware? →

Registry record as of 2026-09-15. Identity, licence and parameter count are properties of a release and do not move; tier can, and the verify link above re-checks it against the live registry rather than this page.