benchgap
Alibaba (Qwen) · model

Qwen3-Coder 480B-A35B benchmark scores

As of 2026-10-29, Qwen3-Coder 480B-A35B (Alibaba (Qwen)) has measured scores on 2 benchmarks and estimated scores on 14 more.

BenchmarkScoreSource
AA-SciCode43.4%estimated ± 3.9 pp, high confidence
SciCode43.1%estimated ± 3.3 pp, high confidence
AA Coding Index39.3%estimated ± 5.3 pp, medium confidence
SWE-bench Verified69.6%measured
SWE-bench Pro46.8%estimated ± 4.6 pp, high confidence
FrontierCode 1.1 Main3.3%estimated ± 2.7 pp, low confidence
Vals LiveCodeBench67.0%estimated ± 9.9 pp, low confidence
Vals SWE-bench62.1%estimated ± 6.8 pp, medium confidence
DeepSWE25.0%estimated ± 12.7 pp, low confidence
SWE Multilingual54.7%measured
Vibe Code Bench0.7%estimated ± 8.1 pp, medium confidence
INCLUDE79.4%estimated ± 2.0 pp, low confidence
NL2Repo25.4%estimated ± 6.2 pp, medium confidence
React Native Evals71.8%estimated ± 2.3 pp, medium confidence
SWE-Rebench57.0%estimated ± 3.6 pp, medium confidence
LiveCodeBench72.7%estimated ± 14.6 pp, low confidence