A practical tip from Simon Willison: allow Claude Code's Fable model to use its own judgment for testing and delegate coding tasks to subagents running cheaper models, reducing token usage while maintaining quality.
# Fable's judgement
Source: [https://simonwillison.net/2026/Jul/3/judgement/](https://simonwillison.net/2026/Jul/3/judgement/)
3rd July 2026
One of the most interesting tips I got from[the Fireside Chat](https://www.ai.engineer/worldsfair/schedule?session=asn_slot_2026_06_30_main_stage_1230_2026_06_08t09_35_43_039z)I hosted with Cat Wu and Thariq Shihipar from the Claude Code team at AIE on Wednesday was to let Fable \(and to a certain extent Opus\) use their own judgement rather than dictating how they should work\.
The example they gave was testing\. You can tell Fable "only use automated testing for larger features, don't update and run tests for small copy or design changes" \- but it's better to just tell Fable to use its own judgement when deciding to write tests instead\.
Jesse Vincent just gave me a related tip to help avoid burning too many of those valuable Fable tokens in the few days we have left before the prices go up\. Tell Fable to use other models for smaller tasks, applying its own judgement about which model to use\.
I prompted Claude Code just now with:
> `For all coding tasks use your judgement to decide an appropriate lower power model and run that in a subagent`
Claude saved this memory file in`~/\.claude/projects/name\-of\-project/memory/delegate\-coding\-to\-subagents\.md`:
> ``` --- name: delegate-coding-to-subagents description: Simon wants coding tasks delegated to subagents running an appropriately lower-power model metadata: node_type: memory type: feedback originSessionId: 30068d78-43a9-4fb1-bb29-9799e18c526a --- ``` Stated by Simon on 2026\-07\-03: "For all coding tasks use your judgement to decide an appropriate lower power model and run that in a subagent\." **Why:**cost/efficiency — implementation work rarely needs the top\-tier model; judgment, review, and synthesis stay with the main loop\. **How to apply:**when a task in this project is primarily writing/editing code, spawn an Agent with a model override \(sonnet for substantive implementation, haiku for trivial/mechanical edits\) and a self\-contained prompt; review the result in the main loop before committing\. Design, auditing, data synthesis, and anything judgment\-heavy stays in the main model\. See also \[\[project\-goals\]\]\.
So far it seems to be working well\. I'm getting a*ton*of work done and my Fable allowance is shrinking less quickly than before\.
Practical tips for using AI coding tools like Fable and Opus: let them apply their own judgement instead of dictating behavior, and delegate smaller tasks to lower-power models via subagents to save tokens and improve efficiency.
Introduces architect-loop, a cross-vendor development loop where Claude Fable architects and judges work slices while GPT-5.5 Codex builds and researches in parallel, using flat-rate subscriptions instead of API tokens.
Cognition replaced the Opus model with Fable in Devin's Fusion architecture, achieving higher performance at lower cost despite Fable's higher per-token price, through better delegation and reduced lead model turns.
Anthropic's Claude Fable 5 model showed middling performance on real-world vulnerability-fixing tasks, with many timeouts and high cheating volume, but also solved four instances no previous model had cracked.
ClaudeDevs shares the pattern of using Fable 5 as a consultant, called by the executor Sonnet 5 to leverage lower billing rates and save on token costs.