Anthropic released Claude Fable 5.2 on November 9, 2026, alongside Claude Mythos 5.2. The new Fable model targets demanding work in software development, mathematics, and science. Its standard API rates are 20% lower than Fable 5.1, with input priced at $8 per million tokens and output at $40 per million tokens. Anthropic’s launch announcement.
What changed in Claude Fable 5.2?
Fable 5.2 supports a 1 million token context window and up to 128,000 output tokens. A large context window can be useful when a task needs several source files, a specification, and test logs together. Anthropic says the model can generate output up to 30% faster than Fable 5.1. Its optional fast mode can speed up token generation by up to 2.5 times. Those figures measure generation speed, while the time needed to finish a coding task also depends on tool calls and tests. Claude Fable 5.2 model overview.
Claude Fable 5.2 API pricing
Here are the standard and fast mode rates in US dollars. Fast mode doubles the listed per-token price for Fable 5.2.
| Model and mode | Input per 1M tokens | Output per 1M tokens |
|---|---|---|
| Fable 5.2 standard | $8 | $40 |
| Fable 5.2 fast | $16 | $80 |
| Fable 5.1 standard | $10 | $50 |
A request with 100,000 uncached input tokens and 10,000 output tokens would cost $1.20 in standard mode or $2.40 in fast mode at those rates. Developers can use claude-fable-5-2 as the model ID on the Claude API. Claude API pricing and migration details.
Who can access Fable 5.2?
Claude subscribers can select Fable 5.2 in the model picker when their plan and workspace settings allow it. The launch offer lets eligible Pro subscribers and standard Team seats spend up to 50% of their weekly usage limit on Fable models through November 16, 2026, at 11:59 p.m. Pacific time. Usage beyond that allowance or after the offer can draw on paid usage credits when enabled. Claude Help Center launch offer.
How Fable 5.2 differs from Mythos 5.2
Both releases share the same underlying model. Fable 5.2 has additional safeguards for cybersecurity and biology tasks. Anthropic restricts Mythos 5.2 access to verified organizations through its specialized programs. Anthropic’s access and safeguards details.
A practical first test for a development team is a reproducible bug that spans several files. Give Fable 5.2 the relevant source files, the failing test, and the expected behavior. Compare the API cost and elapsed time with your current model, then run the tests against the proposed fix.
Posting Komentar