Anthropic shipped fable 5.1 on monday.
It set a new state-of-the-art for coding and cut cache costs 75%. Most of you haven't switched yet. what changed: → CursorBench: 73.4% vs Opus 5's 70.0% — Fable is now the best coding model Anthropic makes → Cache reads dropped from $1
GPT-6 Astra vs. GPT-5.6 Sol (max) vs. Opus 5 vs. Kimi K3 vs. Qwen3.8-Max
Opus 5 Sub agents being released in my codebase
GitHub just dropped Project HydraFusion inside Copilot CLI.
One picker. Multiple models. Runtime picks the workflow — Single, Cascade, or Critique. Vs Claude Opus 5 (GitHub offline evals): • TerminalBench 2.1: +4.9 pts quality, 67% cheaper • DeepSWE: near-parity, 36% cheaper
You have reached the end of the archive
All of Claude Opus 5