Fable 5 is back — best practices for using it well

Quick summary

• Fable 5 is back (July 1) — in-plan on Max and Team Premium, on metered usage credits for Pro.
• Front-load genuinely heavy jobs now, while access is cheapest.
• Record source cards, assumptions, files changed, and verification steps so switching models is easy.
• For routine work, a cheaper model like Sonnet 5 or Opus 4.8 is often enough.

For coding work

Keep the original problem statement, repo state, files changed, tests, and review notes. Run the jobs that genuinely need Fable 5 with caching and batch to blunt the credit rate, and drop to Opus 4.8 or Sonnet 5 for routine iteration.

For research work

Separate source facts from interpretation. Fable's value is strongest where long context and judgment matter — exactly the tasks still worth premium credits — budget them deliberately now that every token is metered.

For content and UI work

Keep screenshots, acceptance criteria, and before/after notes. A strong model's design value is easier to evaluate when the visual target is explicit.

From the Field Guide

Stop guessing which model to use

Weekly verified workflows for the credits era: which model in which seat, what it costs per task, and copy-ready prompts — plus an 8-runbook Starter Pack you can download once and keep. Every claim checked against published pricing, sources shown.