Anthropic released Claude Fable 5.2 on November 9 and cut its API list price by 20% compared with Fable 5.1: $8 per million input tokens and $40 per million output tokens, down from $10 and $50. The prices are reported by eesel, which cites Anthropic’s pricing page, and by CloudZero. Anthropic’s public pricing page did not show per-token rates when we opened it, so confirm the figures in your Claude Platform console before budgeting.
The new price list
Both outlets give the same rates for Fable 5.2:
- Standard: $8 input and $40 output per million tokens (Fable 5.1: $10 and $50).
- Cache: reads $0.20 per million; writes $10 (5 minutes) or $16 (1 hour). Fable 5.1 charged $0.25, $12.50 and $20.
- Batch: $4 input and $20 output, half the standard rate (Fable 5.1: $5 and $25).
- Fast mode: $16 and $80, double standard, with output up to 2.5 times faster. It cannot be combined with Batch.
- Extras: US-only inference bills at 1.1 times the standard rate, and web search costs $10 per 1,000 searches, per CloudZero.
Context is 1 million tokens at the same rate for every request size, with a maximum output of 128,000 tokens, per both outlets. eesel lists the model ID as claude-fable-5-2 on the Claude API, Google Cloud, Microsoft Foundry and Claude Platform on AWS, and anthropic.claude-fable-5-2 on Amazon Bedrock.
Cheaper per token is not always cheaper per task
CloudZero cautions that cost per task depends on the effort level and, citing Artificial Analysis, says Fable 5.2 at maximum effort costs roughly double Fable 5.1 per task. That is a third-party measurement relayed by one outlet, but it is a useful reminder: a 20% lower rate can still produce a higher bill if the model uses more tokens to think. Run your own workload before switching.
Mythos 5.2
Anthropic released Mythos 5.2 alongside it at the same rates, according to both outlets. CloudZero says Mythos is restricted to organisations verified through Anthropic’s programs while Fable 5.2 is generally available. CIO Africa adds that Mythos retains cybersecurity and biological capabilities that are restricted in Fable.
What it means on a subscription plan
eesel, summarising Anthropic’s support article, says Pro and standard Team seats use Fable through usage credits only, while Max, Ultra and premium Team and Enterprise seats include Fable up to 50% of weekly limits. It also reports a launch promotion running November 9 to 16, 2026, that lets Pro and standard Team seats use up to 50% of weekly limits on Fable at no extra cost. Plan terms change, so read Anthropic’s support page before relying on this.
Performance claims
CIO Africa reports Anthropic’s own claims of up to 30% faster responses than Fable 5.1 and a 49.1% score on the FrontierCode 2.0 coding test. These are Anthropic’s figures and have not been independently reproduced.
What to do
- Developers on Fable 5.1: test 5.2 on a sample of real tasks and compare total cost, not just the rate.
- Heavy batch users: the batch rate is now $4 and $20.
- Subscribers: check whether your plan covers Fable or needs usage credits.
Sources
- eesel: Claude Fable 5.2 pricing
- CloudZero: Claude Fable 5.2 and Mythos 5.2 pricing
- CIO Africa: OpenAI, Anthropic bring AI deeper into cloud marketplaces
Image: Anthropic’s San Francisco office, photo by the UK Department for Science, Innovation and Technology, CC BY 2.0, via Wikimedia Commons.
Like our coverage? Add stepphase.com as a preferred source on Google to see more of our stories in Top Stories.
