Across 24 shared benchmarks, Claude Opus 4.8 scores higher on 3 and GPT-6 Astra on 21. The widest gap is Terminal-Bench 4.0, where GPT-6 Astra scores 57.7 against 23.6. Claude Opus 4.8 is the cheaper of the two on tracked API pricing ($5.00 against $10.00 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 (anthropic-official, direct), otherwise the lowest tracked offer.