Claude Sonnet 5.5 xHigh hits #3 in Code Arena WebDev, 2 pts behind GPT-6 Astra at 80% of the price
arena · x · 2026-10-02
LMArena reports Claude Sonnet 5.5 (xHigh reasoning) landed in Code Arena: WebDev at #3 with 1786 pts — just 2 points behind GPT-6 Astra (#2, 1788) at 80% of the price, staying on the Pareto frontier at a blended $8/M tokens.
The earlier High variant scored 1699 (#4), a +159 pt jump over Sonnet 5 (High, 1540), with Reference-Based Design and Simulations leaping from #37 to #4. Domain ranks for xHigh: #2 in Gaming, Reference-Based Design, and Brand & Marketing; #3 in Simulations; #4 in Content Creation Tools and Consumer Product; #6 in Data & Analytics.
Related event: Claude Sonnet 5.5 xHigh Debuts Third on Code Arena(2 posts)→
More from Models
- PostTrainBench v1.2 ranks Fable 5.1 > Opus 5.5 > GPT-6 Astra, adds Harbor support — maksym_andr · 2026-10-02
- Speculation swirls: Gemini 4 and GPT-6 both use looped architectures — PandaAshwinee · 2026-10-02
- Grok Bot turns proactive: pings user mid-flight to fix stale Uber reservation — beffjezos · 2026-10-02
- EMNLP Findings paper: explicit reasoning hurts pointwise rerankers, disabling it helps — lintool · 2026-10-02
- User calls Claude 4.7 'neurotic': it challenges you over the slightest risk — repligate · 2026-10-02
- JevBench to add evals for LLM routing, RAG retrieval, and moderation use cases — airesearch12 · 2026-10-02