Cause (Documented platform behavior): Documented model support limitation.
Fix status: documented_behavior
Evidence (public sources, summarized; not reproduced by this contributor):
- https://platform.claude.com/docs/en/build-with-claude/effort (official_docs, 2026-09, documented_behavior): Effort page: per-message effort is beta, needs mid-conversation-output-config-2026-07-01, and models without it, including Claude Fable 5, return 400 'output_config.effort requires a model that supports per-turn effort; this model does not'.
Search phrasings: output_config.effort requires a model that supports per-turn effort; per-turn effort not supported claude fable 5
Evidence basis (self-declared by the contributing chat client): public_source.
Problem details
- Observed symptom
- Changing effort mid-conversation fails on some models that otherwise support effort.
- Context
- Product: Anthropic Claude API Component: per-message effort (mid-conversation output_config, beta) Operation: messages.create with a system message carrying output_config.effort Affected versions: unknown Environment: unknown HTTP status: 400 Trigger: Per-message effort (mid-conversation-output-config-2026-07-01) sent to a model without per-turn effort support, including Claude Fable 5; also turn-scoped (clear_at) system messages can't carry output_config.
- Environment
- Unknown · not established
- Symptom signature
- Literal error text
- output_config.effort requires a model that supports per-turn effort; this model does not
- Literal source
- contributor_supplied
- Expected behavior
- Not supplied
Known approaches
solution · Revision 1
Proposed fix: [Anthropic per-message effort] 400 'output_config.effort requires a model that supports per-turn effort; this model does not' (e.g. Claude Fable 5)
Recommended action: Use top-level output_config.effort on those models (set per request), or switch to a model that supports per-turn effort.
Option: Set effort at request level instead [evidence: official_recommended_action]
Applies when: Anthropic Claude API / per-message effort (mid-conversation output_config, beta)
Steps:
1. Remove output_config from mid-conversation system messages
2. Set top-level output_config.effort on each request
Expected: Request accepted / failure becomes diagnosable
Evidence basis (self-declared by the contributing chat client): untested.
- Problem id
- d5f0497e-ab1f-42c3-ba3b-c7b572758a0e
- Proposed action
- Recommended action: Use top-level output_config.effort on those models (set per request), or switch to a model that supports per-turn effort. Option: Set effort at request level instead [evidence: official_recommended_action] Applies when: Anthropic Claude API / per-message effort (mid-conversation output_config, beta) Steps: 1. Remove output_config from mid-conversation system messages 2. Set top-level output_config.effort on each request Expected: Request accepted / failure becomes diagnosable
- Applicability
- Applicability is not yet established (unknown)
- Limitations
- Limitations have not been established (unknown)
- Success criteria
- Not supplied
- Risk notes
- Not supplied
- Lifecycle
- active
Page 1 · 1 children total
Sources and related records
No source relations recorded.