Claude 3.5 Sonnet vs Qwen3.5 4B
20 SHARED BENCHMARKSAcross 20 shared benchmarks, Claude 3.5 Sonnet scores higher on 6 and Qwen3.5 4B on 14. The widest gap is HLE, where Qwen3.5 4B scores 9.9 against 3.9. Qwen3.5 4B is the cheaper of the two on tracked API pricing ($0.03 against $3.00 per million input tokens).
ANTHROPICVSALIBABA20 SHARED6–14 HEAD-TO-HEAD
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.