// AI Platform Drift

Claude Fable 5 tokenizer produces roughly 30% more tokens than pre-Opus 4.7 models

Claude Platform release notes · published 2026-11-11 05:30 UTC
ACTION NEEDED BUDGET MODELPRICINGAPI
WHAT CHANGEDClaude 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 content and workload shape. The API now returns thinking: {"type": "disabled"} as unsupported (400 error) on these models; adaptive thinking is the only mode. Requests refused before output generation are not billed. A new stop_reason category "reasoning_extraction" is returned when requests are blocked under Terms of Service restrictions on reverse engineering or duplicating model outputs.
WHY IT MATTERSExisting prompts will consume 30% more input tokens on Claude Fable 5 and Claude Mythos 5, increasing per-request costs proportionally for the same workload. Code that sets thinking: {"type": "disabled"} will receive 400 errors instead of running.
WHAT TO DOUse the token counting API with model: "claude-fable-5" to measure your prompts under the new tokenizer and reforecast costs. Remove thinking: {"type": "disabled"} from requests to Claude Fable 5 and Claude Mythos 5 before migrating.
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
 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 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 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. See Token counts on Claude Fable and Claude Mythos models.
 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).
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 tokenizer breaking change, billing implications, and API incompatibilities require engineering attention.
signalsnew model released (claude-fable-5, claude-mythos-5) · tokenizer change with quantified impact (30% more tokens) · breaking API change (thinking: disabled not supported, returns 400) · billing change (no charge for refused requests before output) · new stop_reason category (reasoning_extraction) · new optional parameter (fallbacks, beta)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)