Across 54 shared benchmarks, DeepSeek-V3 scores higher on 8 and Kimi K2 Instruct on 45, with 1 level. The widest gap is vectara_hallucination_rate, where DeepSeek-V3 scores 6.1 against 17.9. DeepSeek-V3 is the cheaper of the two on tracked API pricing ($0.27 against $0.60 per million input tokens).
Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, direct), otherwise the lowest tracked offer.