LATESTSAAS FOR PEOPLE WHO BUILDAI · SOFTWARE · CREATORS
AI BENCHMARK / ANTHROPIC

Claude Fable 5.2: price, features and how it is used

Claude Fable 5.2 is Anthropic’s model for the hardest reasoning tasks and long agentic workflows. It is released on November 9, 2026 together with Claude Mythos 5.2. It has a context window of 1M tokens and a maximum output of 128K tokens.

Intelligence Indexn/ahigher is better
Input price$8per 1M tokens
Output price$40per 1M tokens
Speedn/atokens per second
TTFTn/aseconds to first token
Context1 milliontokens

Provider: Anthropic · Released: November 2026

Results by test

TestResultSource
FrontierCode 2.0 (xhigh effort)49,1%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
FrontierSWE v269,9%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
AutomationBench51,4%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
Terminal-Bench-Science 0.174,6%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
OSWorld 2.1 (offline set, partial)89,3%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
Humanity's Last Exam (with tools)74,2%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
Chartography (no tools)73,4%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
Remote Labor Index (high effort)41,3%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2
WANDR (relaxed score)82,1%Anthropic, announcement of Claude Fable 5.2 and Mythos 5.2

Intelligence, price and speed by effort level

There is no data yet for this model on effort level, intelligence and price per task.

Data on the Intelligence Index, capability indexes, price per task, speed and latency: Artificial Analysis. Prices per 1M tokens are from official vendor data.

Compare Claude Fable 5.2 with another model

Choose a model and see prices, context and results side by side.

How to use Claude Fable 5.2

Claude Fable 5.2 is used via the Claude Platform (API), via the Claude apps and via the AWS, Google Cloud and Microsoft Azure clouds. The quickest route for a developer is an API key from the Claude Platform, and in the apps the model is available to all users.

  1. Create an account on the Claude Platform and generate an API key. If you do not write code, open the Claude app.
  2. In the request to the Messages API, set the model to claude-fable-5-2, exactly as in the documentation.
  3. Thinking depth is set with output_config.effort. The levels are max, xhigh, high, medium and low, and the API default is high.
  4. Adaptive thinking is always on and cannot be turned off, so effort is your main lever. The documentation recommends starting with high, xhigh for hard tasks, and max only to measure the ceiling.
  5. At high levels, set a large max_tokens, because thinking is counted in it.
  6. For non-English text, keep in mind that non-English text uses more tokens than English, so the same text costs more. Check the quality of answers on your own real examples.

What it is suited for

  • Positioned by Anthropic for the hardest reasoning and long-running agent work.
  • 1M token context window at the standard price.
  • Highest results among Anthropic models in the published table for OSWorld 2.1 and Humanity's Last Exam.

What to keep in mind

  • Adaptive thinking is always on and cannot be turned off. Depth is set with the effort parameter (default high), with no published per-level numbers for tokens, price and time.
  • Forcing tool use (tool_choice any and tool) returns a 400 error.
  • Requires 30-day data retention and is not available with zero data retention without explicit permission from Anthropic.
  • The results are from the provider itself and have safeguards enabled, unless stated otherwise.

Frequently asked questions about Claude Fable 5.2

How much does Claude Fable 5.2 cost?
Via the API it is 8 USD per 1M input and 40 USD per 1M output tokens. There is also a fast mode at double the price, 16 and 80 USD. In the Claude apps the model is included up to 50% of the weekly limits on Pro plans and standard Team seats.
Is Claude Fable 5.2 free?
Not through the API, where you pay per token. The documentation says it is available to all users of the apps, but it does not specify the free plan limits, so check in your own account.
How do you use Claude Fable 5.2 in your own language?
Write your prompt directly in the language you plan to use and compare the answer with what you need. Tokens in non-English languages are more numerous, so the same text costs more than in English. So run a few of your real tasks before using it at scale.
How does Claude Fable 5.2 differ from Claude Opus 5.5?
Fable costs twice as much: 8 and 40 dollars versus 4 and 20 for Opus 5.5. In both, thinking is always on, but the default effort in the API is high for Fable and medium for Opus.

More models in AI Benchmark

Intelligence Index, token, TTFT and effort are explained in the AI Benchmark glossary. Prices are per 1M tokens and may change, so check the provider’s official page.

Sources: Anthropic, model overview · Anthropic, pricing · Anthropic, Fable 5.2 page · Anthropic, announcement of Fable 5.2 and Mythos 5.2

Last checked: 18.11.2026

Compare with other models

Set Claude Fable 5.2: price, features and how it is used to another model by price, context and benchmarks.

Compare

All models by Anthropic: prices and how to use it Claude

FOR CREATORS

Give your product a spot.

Submit the product for editorial review. It will be published after approval.

Submit only public product information. How we handle requests.