List your company

Climate Question Answering on CLIMAQA (held-out)

46.3DBSDataChef-32B
25.18830.66936.1541.631Feb 11, 2026

Evaluation results

MethodLinks
DataChef-32BOracle Upper Bound=true, Samples=322026.0246.3—
Gemini-3-Pro2026.0244.358.4
EXPERTModel=Qwen3-1.7B optimized2026.0244.2—
SOURCEbest2026.0243.6—
Qwen3-Next ⊕ Kimi-K2Reasoning backbone=Qwen3-Next-80B, Coding backbone=Kimi-K2-Instruct2026.0242.641.5
DataChef-32B2026.0242.157.3
Kimi-K22026.0241.818.3
Qwen3-32B2026.0235.620.6
SOURCEavg2026.0226—