Code Arena: GPT-6 Astra ranks #1, but Claude Fable 5.1 wins more head-to-head battles
arena · x · 2026-09-17
LMArena's Code Arena: WebDev leaderboard now ranks OpenAI's GPT-6 Astra (Max) #1 at 1800 points, with Anthropic's Claude Fable 5.1 (Max) second at 1758.
- Win rate measures user preference in anonymous head-to-head battles. Astra clearly wins most matchups, including 72.5% against GPT-5.6 Sol (xHigh).
- The exception is Claude Fable 5.1: in direct battles, Fable 5.1 wins 43.5%, Astra wins 29.0%, and 27.4% end in ties.
- Despite Astra's top ranking, Arena users actually prefer Fable 5.1 when comparing the two models' outputs side by side.
Related event: Code Arena Update: GPT-6 Astra Tops Chart, Win Rates Incoming(2 posts)→
More from Models
- Cloudflare Launches First Stealth Model Union Alpha, Blending Multiple LLMs Per Request — ritakozlov · 2026-09-17
- Union Alpha Allegedly Leaks Its Chain-of-Thought in a Single Request — zainhas · 2026-09-17
- Qwen3.8-Flash-Next on SGLang NVFP4: 254K-Context TTFT Drops 35s→22s, Full Gauntlet Tested — FantasticNature7590 · 2026-09-17
- One audio encoder + LLM serving offline and streaming ASR without accuracy tax — rohanpaul_ai · 2026-09-17
- NetEase Youdao open-sources Confucius4-R2T2, append-only streaming ASR built on Qwen3 — rohanpaul_ai · 2026-09-17
- Claude credit tiers: 4.1B credits for $6/month, 82B for $100 — how many tokens? — MaziyarPanahi · 2026-09-17