// AI Platform Drift

Claude Fable 5 tokenizer produces 30% more tokens, requires 30-day retention

Claude Platform release notes · published 2026-10-10 12:36 UTC
ACTION NEEDED BUDGET MODELPRICINGAPITERMS
WHAT CHANGEDClaude Fable 5 and Claude Mythos 5 launched 9 June 2026 with a new tokenizer that produces roughly 30% more tokens for the same text compared to models before Claude Opus 4.7. Both models support 1M token context windows and 128k max output tokens. Claude Fable 5 requires 30-day data retention; zero data retention is not available. The thinking: {"type": "disabled"} parameter returns a 400 error on both models. A new stop_details.category value "reasoning_extraction" is returned when requests are blocked for reverse engineering attempts.
WHY IT MATTERSThe 30% token increase directly raises your API costs for the same workload. If you are under zero data retention today, you cannot use Claude Fable 5 without changing your data handling agreement. The disabled thinking mode returning 400 will break any code that explicitly disables thinking.
WHAT TO DORun the token counting API with model: "claude-fable-5" on representative prompts to measure the actual token increase for your workload and update cost forecasts. If you use thinking: {"type": "disabled"}, remove that parameter before migrating to Claude Fable 5. If you require zero data retention, confirm whether the 30-day requirement is acceptable before planning any migration.
Model-reported confidence 95% (not measured accuracy) Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 https://platform.claude.com/docs/en/release-notes/overview#june-9-2026
 Tue, 09 Jun 2026 00:00:00 GMT
 
-We've launched Claude Fable 5 (claude-fable-5), our most capable model open to all customers, alongside Claude Mythos 5 (claude-mythos-5) for Project Glasswing participants. Both models support a 1M token context window by default, 128k max output tokens, and always-on adaptive thinking. See Introducing Claude Fable 5 and Claude Mythos 5 for capabilities, API changes, and availability.
+We've launched Claude Fable 5 (claude-fable-5), our most capable model open to all customers, alongside Claude Mythos 5 (claude-mythos-5) for Project Glasswing participants. Both models support a 1M token context window by default, 128k max output tokens, and always-on adaptive thinking. See Claude Fable 5 for specifications, pricing, and availability.
 Claude Fable 5 and Claude Mythos 5 use the tokenizer introduced with Claude Opus 4.7. Compared to models before Claude Opus 4.7, the same text produces roughly 30% more tokens. The exact increase depends on the content and workload shape. Use the token counting API with model: "claude-fable-5" to measure your prompts under the new tokenizer.
 Claude Fable 5 runs safety classifiers on requests and during response generation. When a classifier declines a request, the Messages API returns stop_reason: "refusal". You are not billed for a request refused before any output is generated. An opt-in fallbacks parameter (in beta on the Claude API and Claude Platform on AWS; not supported on the Message Batches API) re-runs refused requests on another model, billed at the fallback model's rates. See Handling stop reasons.
 The stop_details.category field on refusal responses now includes "reasoning_extraction" on Claude Fable 5, returned when a request is blocked under Anthropic's Terms of Service restrictions on reverse engineering or duplicating model outputs. The existing "cyber" and "bio" categories are unchanged. No beta header is required.
-On Claude Fable 5 and Claude Mythos 5, adaptive thinking is the only thinking mode: thinking: {"type": "disabled"} is not supported, and manual extended thinking budgets and assistant prefill are not supported (both return a 400 error). See Migrating from Claude Mythos Preview to Claude Mythos 5.
+On Claude Fable 5 and Claude Mythos 5, adaptive thinking is the only thinking mode: thinking: {"type": "disabled"} is not supported, and manual extended thinking budgets and assistant prefill are not supported (both return a 400 error).
 On Claude Fable 5 and Claude Mythos 5, thinking.display defaults to "omitted", the same as Claude Opus 4.8, Claude Opus 4.7, and Claude Mythos Preview; set display: "summarized" to receive readable thinking summaries. The raw chain of thought is never returned; pass thinking blocks back unchanged in multi-turn conversations on the same model. See Thinking output on Claude Fable 5 and Claude Mythos 5.
 Claude Fable 5 requires 30-day data retention and is not available under zero data retention. See Model-specific data retention requirements.
 Claude Managed Agents now supports scheduled deployments, letting you run sessions on a cron schedule without managing your own scheduler.
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
triage verdictMATERIAL at 0.92, needed 0.45
triage reasoningNew models with specified limits, breaking API changes, tokenizer impact, data retention requirement, and new features.
signalsnew model released (claude-fable-5, claude-mythos-5) · context window specified (1M tokens) · max output tokens specified (128k) · tokenizer change with quantified impact (30% more tokens) · new API field added (stop_details.category: reasoning_extraction) · breaking API constraint (thinking: disabled not supported, returns 400) · data retention requirement introduced (30-day mandatory, zero retention unavailable) · new feature (Claude Managed Agents scheduled deployments) · documentation link changed (reference doc title altered)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)