benchgap
Other · model

Composer 2 benchmark scores

As of 2026-11-21, Composer 2 (Other) has measured scores on 4 benchmarks and estimated scores on 9 more.

BenchmarkScoreSource
AA-SciCode56.6%estimated ± 8.6 pp, low confidence
AA Coding Index89.7%estimated ± 11.1 pp, low confidence
SWE-bench Verified75.7%estimated ± 4.8 pp, medium confidence
Terminal-Bench Hard45.3%estimated ± 4.0 pp, high confidence
Vals SWE-bench100.0%estimated ± 7.0 pp, low confidence
Terminal-Bench 2.1 (Vals AI)54.5%estimated ± 6.4 pp, medium confidence
SWE Multilingual73.7%measured
Vibe Code Bench66.5%estimated ± 6.5 pp, low confidence
INCLUDE84.7%estimated ± 2.0 pp, medium confidence
MMLU-ProX84.4%estimated ± 1.3 pp, high confidence
Terminal-Bench 2.061.7%measured
React Native Evals96.1%measured
SWE-Rebench58.0%measured