We use cookies to enhance your experience, provide personalised content, and analyse our traffic; see our Cookie Policy.
anthropic/claude-fable-5-2The hardest work, carried to the end
Claude Fable 5.2 is the model Anthropic reserves for demanding reasoning and long horizon agentic work: agent sessions that run for hours, multistep research, and analysis carried through to a finished document, spreadsheet or deck. It holds a 1M token context and always thinks before it answers.
Released on 9 November 2026, it replaces Fable 5.1 at a lower price, and Anthropic's own advice is to reach for it when your tests on Opus 5.5 at higher effort still fall short.
Through Tchavi you use it with mobile money, no card and no KYC, with the same API key as 100+ models. Try Opus 5.5 first: if Fable makes the difference, just change the model field.
The model Anthropic reserves for its most demanding work
Agent sessions that run for hours without losing the thread
Reasons over the whole window and connects distant details
From a first question to a finished document, sheet or deck
Breaks a question down, gathers evidence and reconciles it
Reads dense charts, filings and tables inside PDFs
Audit this codebase for security issues, verify each finding against the actual code path, and drop anything you cannot prove is reachable. End with a table of severity, file, evidence.
Reconcile these four quarterly spreadsheets, explain every discrepancy over 1%, and tell me which figures I should not trust and why.
Start with Opus 5.5, as Anthropic recommends. Move to Fable 5.2 for the most demanding reasoning and agent runs measured in hours, or when your tests on Opus 5.5 at high effort still fall short.
Same 1M context and 128K output, and a lower price: Anthropic moved the rate from 10 and 50 dollars per million tokens to 8 and 40. Cached reads stay at 2.5% of the input price. Default effort is high.
1M tokens in and up to 128K out, with images and PDFs in the same request.
Fable 5.2 rejects tool_choice "any" or a named tool with a 400 error. Keep "auto" and say in the prompt when to use the tool, or mark the tool strict so its arguments still validate.
With the OpenAI SDK, change the base URL, use your sk-tch-… key and model: "claude-fable-5-2" on /v1/chat/completions. The Anthropic format /v1/messages endpoint works too.
Yes. Top up your Tchavi account with Orange Money, MTN MoMo, M-Pesa or another mobile money operator. No bank card is needed. Credits land as soon as the payment is confirmed and are deducted as you use them.
Yes. Keep your current OpenAI SDK, just change the base URL to Tchavi's (shown in the API tab) and use your Tchavi key. The rest of your code stays the same.
import Tchavi from '@tchavi/sdk';
const client = new Tchavi({ apiKey: 'API_KEY' });
const response = await client.chat.completions.create({
model: 'claude-fable-5-2',
messages: [
{
"role": "user",
"content": "Hello!"
}
],
});
console.log(response.choices[0].message.content);
console.log('Credits used:', response.tchavi.credits_used);Every parameter this model accepts on the API. Check the SDK for full typings.
modelRequiredID of the chat model (e.g. `claude-fable-5-2`).
messagesRequiredConversation messages. Each item has a `role` (system/user/assistant) and a `content` that's either a plain string or an array of content blocks (`text`, `image_url`, `document_url`).
temperature1Sampling temperature between 0 and 2. Higher values = more random.
max_tokensMaximum number of tokens to generate in the response.
streamfalseIf true, partial message deltas are streamed as Server-Sent Events.
toolsFunction-calling toolbox. Pass an array of `{ type: "function", function: { name, description, parameters } }`. Returned tool calls land in `choices[].message.tool_calls`. Only models with `supportsTools: true` accept this.
tool_choiceHow the model picks among tools: `"auto"` (default), `"none"` to force a plain text reply, or `{ type: "function", function: { name } }` to force a specific tool. Only with `supportsTools: true`.
top_p1Nucleus sampling: only tokens in the top `top_p` probability mass are considered.
stopUp to 4 sequences where the API will stop generating further tokens.
frequency_penalty0Between -2.0 and 2.0. Positive values penalize tokens that have appeared frequently.
presence_penalty0Between -2.0 and 2.0. Positive values encourage the model to talk about new topics.
messages[].content[].image_urlAttach an image on a user message. Pass a content block `{ "type": "image_url", "image_url": { "url": ... } }`. PNG / JPEG / WebP, up to 4 per request. Only models with `supportsImageInput: true` accept this block (e.g. `gpt-4o`, `claude-opus-4-7`).
messages[].content[].document_urlAttach a PDF on a user message. Pass a content block `{ "type": "document_url", "document_url": { "url": ... } }`. Up to 5 PDFs per request. Only models with `supportsDocuments: true` accept this block (e.g. `claude-sonnet-4-6`, `gpt-4.1`).
effort"high"Reasoning depth on Anthropic adaptive-thinking models (Opus 4.7+ / Sonnet 4.6+). Values `"low"`, `"medium"`, `"high"`. Higher = more deliberate, slower. Defaults to `"high"` on Opus 4.8.
Estimated over 30 days, for a typical message: ~500 input + ~150 output tokens, billed to the thousandth of a credit.