Anthropic's rules of thumb: start on Opus 5.5, escalate to Fable 5.2 for the hardest work
JeremyNguyenPhD · x · 2026-11-10
Anthropic's Claude developer blog shares rules of thumb for choosing between models:
- Start most work on Opus 5.5 at default effort (medium): it's the Claude Code default and costs half as much per token as Fable 5.2 — recommended for most coding and knowledge work.
- Move down to Sonnet 5.5 for well-scoped everyday work like bug fixes or spec-driven features; but clear specs don't always mean Sonnet — migrations or performance optimizations may still warrant Opus 5.5 or Fable 5.2.
- Move down to Haiku 5.5 for the simplest high-volume work (classification, routing, extraction), and as subagents handling side tasks for larger models.
- Move up to Fable 5.2 for the hardest coding, finance and science projects, when Opus 5.5 at high effort falls short, or for long unattended runs.
Related event: Anthropic Releases Official Guide on Choosing Claude Models(2 posts)→
More from coding & agent
- Ex-Zalando AI lead's thread: agentic AI changes what a company is, not just add-ons — ilparone · 2026-11-10
- PSAISuite Gives Claude Fable-5.2 Real PowerShell Hands for Agent Loops — dfinke · 2026-11-10
- Coding Assistants Just Moved Your Bottleneck: 27 PRs Stuck Waiting on Review — Pavan_Belagatti · 2026-11-10
- AIMUG meetup: with agents you're the manager—define work, context and tools — colinmcnamara · 2026-11-10
- Chris Albon jokes Codex should be your default browser — chrisalbon · 2026-11-10
- Fable 5.2 built an interactive 3D globe of Earth's history from one prompt, ~$550 in tokens — Chris_Brannigan · 2026-11-10