Claude Fable 5.1

Claude Fable 5.1 (anthropic/claude-fable-5.1) is Anthropic’s most capable widely released model and the successor to Fable 5: agentic coding, long-running agentic workflows and knowledge work. Besides the OpenAI-compatible POST /v1/chat/completions, the model is available through the Anthropic-protocol POST /v1/messages — it works in Claude Code and the Anthropic SDK directly.

Parameters and capabilities

Capability	Value
model	anthropic/claude-fable-5.1 (or short claude-fable-5.1)
Image input	Yes — image_url with a data: URL in the last message
File input	Yes — PDF, DOCX, TXT, CSV, XLSX, PPTX, up to 5 per request
Reasoning	reasoning_effort: low, medium, high, xhigh, max; disabling is not supported — reasoning is built into the model
Context cache read	27.00 ₽ ($0.30) per 1M tokens
Context cache write	1348.60 ₽ ($15) per 1M tokens
Context / output	1,000,000 tokens / up to 128,000

A cache read on Fable 5.1 costs one fortieth of the input price (27.00 ₽ vs 1078.90 ₽) — noticeably cheaper than the Claude-family average; a write is ×1.25. Reasoning is billed as output tokens and consumes the max_tokens budget — see Context caching for details.

Examples

A request with a reasoning depth

Python

from openai import OpenAI

client = OpenAI(base_url="https://api.mixen.ai/v1", api_key="mxn-...")

resp = client.chat.completions.create(
    model="anthropic/claude-fable-5.1",
    messages=[{"role": "user", "content": "Refactor this FastAPI service: split the monolith into modules and justify the boundaries"}],
    reasoning_effort="high",
)
print(resp.choices[0].message.content)

cURL

curl https://api.mixen.ai/v1/chat/completions \
  -H "Authorization: Bearer $MIXEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-fable-5.1",
    "messages": [{"role": "user", "content": "Refactor this FastAPI service: split the monolith into modules and justify the boundaries"}],
    "reasoning_effort": "high"
  }'

Streaming in Python

stream = client.chat.completions.create(
    model="anthropic/claude-fable-5.1",
    messages=[{"role": "user", "content": "Plan a zero-downtime database migration: steps, risks, rollbacks"}],
    stream=True,
    stream_options={"include_usage": True},
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Anthropic SDK

The same key and balance work via POST /v1/messages. The base URL has no /v1 suffix: the SDK appends it to the path itself, unlike the OpenAI SDK.

from anthropic import AsyncAnthropic

client = AsyncAnthropic(
    base_url="https://api.mixen.ai",
    api_key="mxn-...",  # the x-api-key header is supported
)
msg = await client.messages.create(
    model="anthropic/claude-fable-5.1",
    max_tokens=2048,
    messages=[{"role": "user", "content": "Hi!"}],
)
print(msg.content[0].text)

In Claude Code the model takes the top slot via ANTHROPIC_DEFAULT_OPUS_MODEL — see the Claude Code guide.

Price

Per 1M tokens: input — 1078.90 ₽ ($12), output — 5394.40 ₽ ($60). Context cache: read — 27.00 ₽ ($0.30), write — 1348.60 ₽ ($15). Showcase prices follow the exchange rate — check current values in the catalog.

Strengths and limits

• Agentic coding, long-running agentic workflows and knowledge work are the model’s profile per Anthropic; especially strong at long code refactors, front-end and visual code generation, finance and analysis tasks.
• More concise than Fable 5 in plans and summaries; Anthropic recommends it as a direct upgrade from Fable 5 and alongside Opus 5 on reasoning-heavy tasks.
• Reasoning is always on: depth is controlled by reasoning_effort up to xhigh and max; the off value is not supported by the model.
• Claude Code and the Anthropic SDK connect directly — tool use, streaming, count_tokens and cache_control work.
• Prompt caching: a cache read is one fortieth of the input price (a Claude-family record), the main saving on long sessions.
• Reasoning consumes max_tokens; up to 150,000 characters are taken from each file, no OCR.
• Output at 5394.40 ₽ per 1M tokens is the top tier; for high-volume routine Claude Haiku 4.5 is cheaper — 107.90 ₽ input / 539.40 ₽ output.

All models — in the catalog. General text workflow — the chat guide.