Claude Fable 5.2 API Pricing and Day 0 Support on LiteLLM

LiteLLM now supports Claude Fable 5.2 on Day 0. Call it through the LiteLLM AI Gateway and track spend, rate limits and logs in one place.

Claude Fable 5.2 pricing​

• Input: $8.00 / MTok
• Output: $40.00 / MTok
• Cache read: $0.20 / MTok
• Cache write: $10.00 / MTok for 5 minutes, $16.00 / MTok for 1 hour
• Batch: $4.00 / MTok input and $20.00 / MTok output

Input and output both cost 20% less than Claude Fable 5.1.

Prices on every provider: https://models.litellm.ai/models

What's new in Claude Fable 5.2​

• 20% cheaper than Fable 5.1 on input and output, with cache reads down from $0.25 to $0.20
• 1M-token context window and up to 128K output tokens
• Adaptive thinking is always on, with five effort levels from low to max and high as the default
• Fast mode is available on the Claude API at $16 / MTok input and $80 / MTok output

Use Claude Fable 5.2 with LiteLLM​

Pricing landed in PR #50715. No upgrade needed: open Price Data under Models + Endpoints in the UI and click Reload Price Data (or POST /reload/model_cost_map), on v1.76.0 and above.

model_list:
  - model_name: claude-fable-5-2
    litellm_params:
      model: anthropic/claude-fable-5-2
      api_key: os.environ/ANTHROPIC_API_KEY

docker run -d \
  -p 4000:4000 \
  -e ANTHROPIC_API_KEY=$ANTHROPIC_API_KEY \
  -v $(pwd)/config.yaml:/app/config.yaml \
  ghcr.io/berriai/litellm:v1.106.2 \
  --config /app/config.yaml

model_list:
  - model_name: claude-fable-5-2
    litellm_params:
      model: bedrock/claude_platform/claude-fable-5-2
      aws_access_key_id: os.environ/AWS_ACCESS_KEY_ID
      aws_secret_access_key: os.environ/AWS_SECRET_ACCESS_KEY
      aws_region_name: us-east-1

docker run -d \
  -p 4000:4000 \
  -e AWS_ACCESS_KEY_ID=$AWS_ACCESS_KEY_ID \
  -e AWS_SECRET_ACCESS_KEY=$AWS_SECRET_ACCESS_KEY \
  -v $(pwd)/config.yaml:/app/config.yaml \
  ghcr.io/berriai/litellm:v1.106.2 \
  --config /app/config.yaml

3. Test it!

curl --location 'http://0.0.0.0:4000/chat/completions' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer $LITELLM_KEY' \
--data '{
  "model": "claude-fable-5-2",
  "messages": [
    {
      "role": "user",
      "content": "what llm are you"
    }
  ]
}'

Anthropic also serves Fable 5.2 on Amazon Bedrock as anthropic.claude-fable-5-2, on Google Cloud and in Microsoft Foundry as claude-fable-5-2. Cost map entries for those routes are in flight in PR #50713 and PR #50707; until one of them merges, requests on those routes bill at $0 in the gateway.

FAQ​

How much does Claude Fable 5.2 cost?​

$8 per 1M input tokens and $40 per 1M output tokens. Cache reads cost $0.20 per 1M tokens and batch requests cost $4 and $20.

Is Claude Fable 5.2 cheaper than Fable 5.1?​

Yes, by 20% on both input and output. Fable 5.1 costs $10 and $50 per 1M tokens.

What is the context window of Claude Fable 5.2?​

1M tokens, with up to 128K output tokens per request.

How do I call Claude Fable 5.2 with an OpenAI-compatible API?​

Add anthropic/claude-fable-5-2 to your LiteLLM config, then send a standard /chat/completions request to the proxy with "model": "claude-fable-5-2".

Which models can read Claude Fable 5.2 thinking blocks?​

Only Fable 5.2 and Mythos 5.2. A fallback onto it keeps the conversation's reasoning; a fallback off it drops it, Opus 5.5 and Fable 5.1 included.

Why does Claude Fable 5.2 return "The block is bound to a different conversation"?​

Editing system, tools or an earlier message after a thinking block trips the prefix check, which now applies to every account.

Does Claude Fable 5.2 support the computer_20251124 tool?​

No. Use computer_toolset_20260801; the earlier tool returns a 400 everywhere except Bedrock.

Can I force a tool call on Claude Fable 5.2?​

No, tool_choice of any or tool returns a 400. LiteLLM maps tool_choice: "required" onto any, so keep it at auto and use strict: true.

Feedback​

Questions and feedback go in GitHub discussions.