Updated · June 11, 2026

Claude Fable 5: Complete Guide

Anthropic just released Claude Fable 5 — the first publicly available Mythos-class model. Here's everything you need to know: how to access it, what it costs, how to use the API, and tips to get the most out of it.

Just released June 9: Claude Fable 5 is now available to all Claude users. It's the #1 model on every major coding benchmark. See full benchmark comparison →

📋 Table of Contents

What is Claude Fable 5? How to Access Claude Fable 5 Pricing — Free vs Pro vs Max vs API API Setup Guide Pro Tips for Getting the Most Out of Fable 5 Fable 5 vs the Competition FAQ
⚠️ Disclosure: Some links on this page are affiliate links. We may earn a commission at no extra cost to you if you sign up through our links. This helps us keep TrendPulse free. We only recommend tools we've actually tested.

What is Claude Fable 5?

Claude Fable 5 is Anthropic's latest and most capable AI model, released on June 9, 2026. It's the first model from Anthropic's Mythos class — previously restricted to select partner institutions due to its advanced capabilities in finding software security vulnerabilities — that's now available to the general public.

Fable 5 is a "safe" version of Mythos with guardrails that prevent misuse while still delivering state-of-the-art performance. It's already #1 on the Artificial Analysis Intelligence Index with a score of 61.4, the first model to break 60.

80.3%
SWE-bench Pro
#1 Overall
95.0%
SWE-bench Verified
Record High
61.4
Intelligence Index
First to break 60
200K
Context Window
~500 pages

How to Access Claude Fable 5

There are three ways to use Claude Fable 5, depending on your needs:

1
Web Interface (claude.ai)
Go to claude.ai and sign in. Free users get limited Fable 5 messages. Pro subscribers ($20/mo) get moderate access. Max subscribers get heavy access. This is the easiest way to start.
Try Claude Free →
2
API (Developers)
Sign up at console.anthropic.com, get an API key, and use the model ID claude-fable-5-20250609. Pay-per-token: $10/1M input, $50/1M output. Best for building apps and integrations.
Get API Key →
3
Claude Code (Terminal)
Install with npm install -g @anthropic-ai/claude-code, then run claude in your project directory. Requires Pro or Max plan. Fable 5 is now the default model for Claude Code.
Try Claude Code →

Pricing — Free vs Pro vs Max vs API

Free
$0
forever
  • Limited Fable 5 messages/day (~5-10)
  • Unlimited Claude Haiku
  • Web interface only
  • No Claude Code access
  • Standard response speed
Sign Up Free
Best Value
Pro
$20
per month
  • Moderate Fable 5 usage (~50-100 msgs/day)
  • Unlimited Haiku + Sonnet
  • Claude Code included
  • Projects & Artifacts
  • Priority access during peak
Get Pro — $20/mo →
Max 5x
$100
per month
  • 5x Pro usage of Fable 5
  • Everything in Pro
  • Extended thinking with Fable 5
  • Priority queue
  • Best for daily heavy use
Get Max 5x →
Max 20x
$200
per month
  • 20x Pro usage of Fable 5
  • Everything in Max 5x
  • For power users & teams
  • Fastest response times
  • Near-unlimited Fable 5
Get Max 20x →

API Pricing (Pay-Per-Use)

ModelInput (/1M tokens)Output (/1M tokens)ContextBest For
Claude Fable 5$10.00$50.00200KComplex coding, research
Claude Opus 4.8$15.00$75.00200KRestricted Mythos tasks
Claude Sonnet 4$3.00$15.00200KDaily work, good balance
Claude Haiku 3.5$0.80$4.00200KHigh-volume, fast tasks

💡 Pro Tip: Prompt Caching Saves Up to 90%

If you're sending similar prompts (e.g., system prompts, long documents), enable prompt caching in the API. Cached input tokens cost only $1/1M for Fable 5 (90% discount). This makes Fable 5 much more affordable for repeated use cases like chatbots, code review tools, and document analysis.

API Setup Guide

1
Get Your API Key
Go to console.anthropic.com, sign up or log in, navigate to API Keys, and create a new key. Store it securely — you won't be able to see it again.
2
Install the SDK
Run pip install anthropic (Python) or npm install @anthropic-ai/sdk (Node.js). The SDK handles authentication and request formatting automatically.
3
Make Your First Call
Use the model ID claude-fable-5-20250609 in your API calls. See the code examples below for Python and Node.js.

Python Example

import anthropic client = anthropic.Anthropic() # Uses ANTHROPIC_API_KEY env var message = client.messages.create( model="claude-fable-5-20250609", max_tokens=4096, messages=[ {"role": "user", "content": "Explain quantum computing in simple terms"} ] ) print(message.content[0].text)

Node.js Example

import Anthropic from '@anthropic-ai/sdk'; const client = new Anthropic(); const message = await client.messages.create({ model: 'claude-fable-5-20250609', max_tokens: 4096, messages: [ { role: 'user', content: 'Explain quantum computing in simple terms' } ] }); console.log(message.content[0].text);

💰 Cost Calculator

A typical conversation (2K input + 1K output tokens) costs about $0.07 with Fable 5. A heavy coding session with 20K input and 10K output costs about $0.70. With prompt caching on system prompts, real-world costs are often 50-70% lower than raw token pricing.

Pro Tips for Getting the Most Out of Fable 5

🎯 Use Extended Thinking for Complex Tasks

Enable extended thinking (budget_tokens parameter) for complex coding, math, and reasoning tasks. Fable 5's thinking is significantly better than previous models. Set budget_tokens to 10000-50000 for hard problems.

📝 Be Specific in Prompts

Fable 5 responds extremely well to detailed instructions. Instead of "write a function," say "write a Python function that takes a list of stock prices and returns the maximum profit from one buy-sell transaction, handling edge cases."

🔗 Use with Claude Code

Claude Code + Fable 5 is the most powerful AI coding setup available. Use it for complex refactors, multi-file edits, and understanding large codebases. The combination scored 80.3% on SWE-bench Pro — far above any competitor.

💾 Enable Prompt Caching

For API users: cache your system prompts and frequently-used context. This reduces costs by up to 90% on repeated calls. Set cache_control: {"type": "ephemeral"} on messages you want to cache.

📚 Upload Long Documents

With 200K context (~500 pages), Fable 5 can analyze entire codebases, legal documents, or research papers in one shot. Upload directly in claude.ai or pass via API. It handles long context better than any previous model.

⚡ Use Sonnet for Simple Tasks

Don't waste Fable 5 tokens on simple tasks. Use Claude Sonnet 4 ($3/1M input) for summarization, translation, and basic Q&A. Reserve Fable 5 for complex reasoning, coding, and analysis where its superiority shines.

Fable 5 vs the Competition

FeatureClaude Fable 5GPT-5.5 (OpenAI)Gemini 3.1 ProDeepSeek V3
ReleaseJune 9, 2026March 2026April 2026Jan 2026
SWE-bench Pro80.3%58.6%54.2%48.7%
Input Price$10/1M$5/1M$3.50/1M$0.27/1M
Output Price$50/1M$25/1M$14/1M$1.10/1M
Context Window200K256K2M128K
Best ForCoding, reasoningGeneral useLong contextBudget-friendly
Free TierLimited Fable 5Limited GPT-5.5Limited 3.1 ProUnlimited

🏆 Our Verdict

For coding and complex reasoning: Claude Fable 5 is the clear winner — no contest.
For budget-conscious users: DeepSeek V3 is unbeatable on price.
For long documents (500+ pages): Gemini 3.1 Pro's 2M context window is unmatched.
For general daily use: GPT-5.5 offers the best balance of capability and cost.

Ready to Try Claude Fable 5?

Start for free, or get Pro for $20/month to unlock full Fable 5 access + Claude Code.

Get Claude Pro — $20/mo → Get API Access →

Frequently Asked Questions

How do I access Claude Fable 5?
Go to claude.ai and sign in. Free users get limited Fable 5 messages per day. For full access, get Pro ($20/mo) or Max ($100-200/mo). Developers can use the API at console.anthropic.com with pay-per-token pricing.
Is Claude Fable 5 free?
Partially. Free users get ~5-10 Fable 5 messages per day, after which the system falls back to Claude Haiku. For unlimited Fable 5 access, you need Pro ($20/mo) or higher. API access is always pay-per-use.
What's the difference between Fable 5 and Mythos 5?
Mythos 5 is Anthropic's unrestricted most-powerful model, available only to select partner institutions. Fable 5 is the public-facing version with safety guardrails. Fable 5 performs close to Opus 4.8 on unrestricted benchmarks but is safe for general use. The full Mythos 5 remains behind access gates.
Can I use Claude Fable 5 with Claude Code?
Yes! Claude Code now uses Fable 5 as its default model. Install with npm install -g @anthropic-ai/claude-code, then run 'claude' in your project directory. Requires a Pro or Max subscription. This is currently the most powerful AI coding setup available.
How much does Claude Fable 5 API cost per request?
A typical conversation (2K input + 1K output tokens) costs about $0.07. A heavy coding session (20K input + 10K output) costs about $0.70. With prompt caching, costs are often 50-70% lower. The API pricing is $10/1M input tokens and $50/1M output tokens.
Is Claude Fable 5 better than GPT-5.5 for coding?
Yes, significantly. Fable 5 scores 80.3% on SWE-bench Pro vs GPT-5.5's 58.6% — that's a 37% relative improvement. On SWE-bench Verified, it's 95.0% vs 78.2%. For any coding task, Fable 5 is the clear winner.
What is the model ID for Claude Fable 5 in the API?
The model ID is claude-fable-5-20250609. Use this in your API calls: client.messages.create(model="claude-fable-5-20250609", ...). You can also use the alias "claude-fable-5" which will always point to the latest version.

Related Guides

Claude Fable 5 vs GPT-5.5 vs Gemini 3.1 Pro

Full benchmark comparison →
Best AI Coding Agents 2026

Claude Code vs Cursor vs Copilot →
Best AI Subscriptions 2026

Which AI plan is worth paying for →
Best Free AI Chatbots 2026

Every free AI chatbot compared →