New Flagship Model Added: Claude Fable 5.1

Date: 2026-09-02 / (1405-06-11)

Summary

Claude Fable 5.1, Anthropic's new flagship for coding, knowledge work, and long-running problem solving, is now available on AvalAI as claude-fable-5-1. It provides a 1 million-token input window, up to 128,000 output tokens, always-on adaptive thinking, and cache reads priced at $0.25 per 1 million tokens.

---

Details

Anthropic

We announce access to Claude Fable 5.1 (claude-fable-5-1), Anthropic's generally available flagship for demanding software engineering, research, knowledge work, computer use, and multi-step agentic workflows. Anthropic reports stronger results than Claude Fable 5 across agentic coding, multidisciplinary reasoning, business workflows, and scientific research, with improved readability during long tasks.

Key features:

• Advanced coding and knowledge work: Designed to find root causes, work across large codebases, and produce complete research and professional deliverables
• Long-horizon agents: Better suited to sustained, tool-heavy workflows that require planning, verification, and reprioritization
• Scientific research: Anthropic reports gains in agentic scientific research, computational analysis, and multidisciplinary reasoning
• 1M-token input window: Up to 1,000,000 input tokens for large repositories, extensive document sets, and long conversations
• 128K output capacity: Up to 128,000 output tokens for substantial code, analysis, and structured deliverables
• Always-on adaptive thinking: Supports configurable effort, including xhigh and max, while keeping reasoning enabled
• Multimodal and developer features: Vision, PDF input, computer use, function calling, tool choice, native structured output, response schemas, and prompt caching
• Lower cache-read cost: Cached input costs $0.25 per 1 million tokens, 75% below Claude Fable 5's $1.00 rate
• Endpoint support: Full support on v1/chat/completions and v1/messages; partial support on v1/responses
• Account access: Available to AvalAI Tier 2 and higher accounts

Endpoint Availability

Endpoint	Support	Notes
v1/chat/completions	Full	OpenAI-compatible chat, reasoning, vision, structured output, and tool workflows
v1/messages	Full	Native Anthropic-compatible messages, adaptive thinking, and tool use
v1/responses	Partial	Verify every required parameter and tool before production use

---

Pricing

Prices are in USD per 1 million tokens.

Model	Input	Cached Input	Cache Creation Input	Output
claude-fable-5-1	$10.00	$0.25	$12.50	$50.00

Input, cache-creation, and output rates remain aligned with Claude Fable 5. The cached-input rate falls from $1.00 to $0.25 per 1 million tokens. Anthropic estimates that this can reduce typical token-billed workload costs by about 25%, with larger savings possible for cache-heavy agentic work; actual savings depend on cache reuse and request composition.

---

API Request and Response Example

Request

curl https://api.avalai.ir/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AVALAI_API_KEY" \
  -d '{
    "model": "claude-fable-5-1",
    "messages": [
      {
        "role": "user",
        "content": "Review this incident timeline, identify the most likely root cause, and propose a verification plan."
      }
    ]
  }'

Response

The shortened response below illustrates the standard Chat Completions structure. Token counts and cost are illustrative and vary with the request.

{
  "id": "chatcmpl-claude-fable-5-1-example",
  "created": 1788350400,
  "model": "claude-fable-5-1",
  "object": "chat.completion",
  "system_fingerprint": null,
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "The strongest hypothesis is a retry race between the queue consumer and the timeout handler. Verify it by correlating duplicate job IDs with lease expiry, reproducing the timing under load, and confirming that an idempotency guard removes the duplicate writes.",
        "role": "assistant",
        "thinking_blocks": [],
        "annotations": []
      }
    }
  ],
  "usage": {
    "completion_tokens": 70,
    "prompt_tokens": 25,
    "total_tokens": 95,
    "completion_tokens_details": null,
    "prompt_tokens_details": {
      "audio_tokens": null,
      "cached_tokens": 0,
      "text_tokens": 25,
      "image_tokens": null
    }
  },
  "estimated_cost": {
    "unit": "0.0037500000",
    "irt": 574.5,
    "exchange_rate": 153200
  }
}

---

SDK Usage Examples

Bash

curl https://api.avalai.ir/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $AVALAI_API_KEY" \
  -d '{
    "model": "claude-fable-5-1",
    "messages": [
      {
        "role": "user",
        "content": "Analyze this repository architecture and propose a verified modernization plan."
      }
    ]
  }'

Python

from openai import OpenAI

client = OpenAI(
    api_key="your-avalai-api-key",
    base_url="https://api.avalai.ir/v1",
)

response = client.chat.completions.create(
    model="claude-fable-5-1",
    messages=[
        {
            "role": "user",
            "content": "Analyze this repository architecture and propose a verified modernization plan.",
        }
    ],
)

print(response.choices[0].message.content)

JavaScript

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.AVALAI_API_KEY,
  baseURL: "https://api.avalai.ir/v1",
});

const response = await client.chat.completions.create({
  model: "claude-fable-5-1",
  messages: [
    {
      role: "user",
      content: "Analyze this repository architecture and propose a verified modernization plan.",
    },
  ],
});

console.log(response.choices[0].message.content);

Native Anthropic SDK (v1/messages)

Bash

curl https://api.avalai.ir/v1/messages \
  -H "Content-Type: application/json" \
  -H "x-api-key: $AVALAI_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -d '{
    "model": "claude-fable-5-1",
    "max_tokens": 2048,
    "messages": [
      {
        "role": "user",
        "content": "Find the root cause of this intermittent concurrency failure and design a regression test."
      }
    ]
  }'

Python

import anthropic

client = anthropic.Anthropic(
    api_key="your-avalai-api-key",
    base_url="https://api.avalai.ir",
)

message = client.messages.create(
    model="claude-fable-5-1",
    max_tokens=2048,
    messages=[
        {
            "role": "user",
            "content": "Find the root cause of this intermittent concurrency failure and design a regression test.",
        }
    ],
)

print(message.content[0].text)

JavaScript

import Anthropic from "@anthropic-ai/sdk";

const client = new Anthropic({
  apiKey: process.env.AVALAI_API_KEY,
  baseURL: "https://api.avalai.ir",
});

const message = await client.messages.create({
  model: "claude-fable-5-1",
  max_tokens: 2048,
  messages: [
    {
      role: "user",
      content: "Find the root cause of this intermittent concurrency failure and design a regression test.",
    },
  ],
});

console.log(message.content[0].text);

---

Adaptive Thinking and Effort

Claude Fable 5.1 keeps thinking enabled and supports adaptive thinking with configurable effort. Start with medium or high, then increase to xhigh or max only when evaluations show a meaningful improvement in task success. Do not ask the model to expose hidden chain-of-thought; request a concise rationale, evidence, or verification checklist instead.

from openai import OpenAI

client = OpenAI(
    api_key="your-avalai-api-key",
    base_url="https://api.avalai.ir/v1",
)

response = client.chat.completions.create(
    model="claude-fable-5-1",
    messages=[
        {
            "role": "user",
            "content": "Design a staged migration and verify its rollback strategy.",
        }
    ],
    extra_body={
        "thinking": {"type": "adaptive"},
        "output_config": {"effort": "high"},
    },
)

print(response.choices[0].message.content)

---

Recommended Workloads

Workload	Why Claude Fable 5.1 Fits
Complex software engineering	Strong root-cause analysis, codebase understanding, verification, and tool use
Long-running agents	Sustains multi-step work, reprioritizes as evidence changes, and checks intermediate results
Knowledge work	Handles research, document analysis, due diligence, numerical reasoning, and structured deliverables
Scientific analysis	Anthropic reports stronger agentic research and multidisciplinary reasoning than Claude Fable 5
Computer use	Supports screenshots, images, PDFs, and browser or desktop workflows
Cache-heavy workflows	The $0.25 cached-input rate reduces repeated-context cost for eligible prompt-cache hits

---

Related Links

• Claude Fable 5.1 Model Details
• Anthropic Models Overview
• Reasoning Models Guide
• Chat Completions API Reference
• Messages API Reference
• Pricing