Claude Fable 5: API, Benchmarks, Pricing & How to Use It

Claude Fable 5.1 is the current version of Anthropic's Mythos-class Fable model, built for long-running coding, research, knowledge work, and agentic workflows. It supports a 1-million-token context window and 128K maximum output, with $10/M input and $50/M output token pricing. Anthropic says Fable 5.1 costs about 25% less than Fable 5 for typical workloads because of cheaper cache reads. 

If you've spent the last year watching models inch forward half a point at a time on a benchmark, Fable 5 is a jolt. It posts 80.3% on SWE-Bench Pro while the next-best model sits 11 points behind. It finished a migration in a 50-million-line codebase in a day. And Andrej Karpathy called it "a major-version-bump-deserving step change forward." This guide covers what it is, how it benchmarks, what it costs, and exactly how to put it into production, with every figure traced back to Anthropic's launch materials.

What is Claude Fable 5?

Claude Fable 5 is Anthropic’s model designed for complex, long-running tasks that require more than a quick answer. It was built for work such as advanced coding, deep research, large-scale analysis, and other agentic tasks where Claude may need to reason through multiple steps and use tools over an extended workflow.

Anthropic positioned Fable 5 as its first Mythos-class model for general use. The idea was simple: the more complicated and time-consuming a task becomes, the more useful a model designed for long-horizon work can be. For a simple question, you may not notice a major difference. For a task involving dozens of steps, large amounts of information, or sustained tool use, the difference becomes more relevant.

However, Claude Fable 5 is no longer the latest version. Anthropic released Claude Fable 5.1 in September 2026, bringing improvements to long-running agentic coding, multi-step research, and work involving documents, spreadsheets, and presentations. Fable 5.1 also supports a 1-million-token context window and up to 128,000 output tokens.

There is also a close connection between Fable and Anthropic’s Mythos models. Fable 5.1 and Mythos 5.1 share the same underlying model, but they are offered with different safeguards. Fable is the version intended for general availability, while Mythos is restricted to vetted organisations working on areas such as cybersecurity and life sciences.

So, when you see Claude Fable 5 mentioned today, it refers to the model that introduced this model family. If you are looking for the current version, Claude Fable 5.1 is the relevant model to explore.

Build with Claude Fable 5.1 on TrueFoundry

Access advanced models through a unified AI Gateway with built-in controls for security, observability, and cost management.

Book a Demo

Key specifications

Claude Fable 5 benchmarks: the full table

Here's the complete benchmark set Anthropic published at launch. A few things to read carefully:

• The scores shown are the higher of Mythos 5 / Fable 5, they're within 1-3 points of each other on most tests.
• The starred (*) rows are the exception: on cybersecurity, biology, and some reasoning tasks, Fable 5's blocking safeguards kick in, so real-world Fable 5 performance on those rows lands closer to Opus 4.8 (more on that below). Don't quote the starred numbers as "Fable 5" without that caveat.

H2: What actually stands out

• Coding is the headline. +11 points over Opus 4.8 on SWE-Bench Pro, and more than 2x on FrontierCode (Diamond), a benchmark for difficult tasks held to production-codebase standards, where Fable 5 leads even at medium effort. That last part matters: it means strong results without burning maximum reasoning tokens.
• Spatial reasoning nearly tripled over Opus 4.8 (38.6% vs 14.5%).
• Legal reasoning went from "barely passing" to category-leading (13.3% vs GPT-5.5's 2.1% and Gemini's 0.0%).
• The Stripe story puts numbers in context: in a 50-million-line Ruby codebase, Fable 5 ran a codebase-wide migration in a single day, work Stripe estimated at over two months for a full team.
• Vision went SOTA. Fable 5 beat Pokémon FireRed using only raw game screenshots, no maps, no navigation aids, no helper harness that earlier models needed.

Andrej Karpathy's launch-day reaction captured the mood, and notably backed up the "step change" framing with a qualitative read, not just the scores:

Claude Fable 5 vs Other AI Models: How Does It Compare?

Claude Fable 5 vs Opus 4.8

Here is the difference between Claude Fable 5 vs Opus 4.8: 

Claude Fable 5 vs GPT-5.5

Claude Fable 5 vs Gemini

Claude Fable 5 vs Mythos 5

What Are the Limitations of Claude Fable 5?

Claude Fable 5 is built for complex, long-running tasks, but it is not the right choice for every workload. Its biggest limitations are its higher cost, potentially high token usage, and safety restrictions that can affect how it handles certain requests.

• Higher cost: Fable 5.1 costs $10 per million input tokens and $50 per million output tokens. That can make it expensive for high-volume workloads, especially when a smaller model can handle the same task.
• High usage on complex tasks: Long coding sessions, research workflows, and agentic tasks can consume a significant number of tokens. This can make Fable less practical when you have strict usage limits or a fixed budget.
• Safety restrictions: Fable has safeguards designed to prevent harmful or high-risk use. In some cases, these safeguards can lead to refusals or require you to rephrase a legitimate request, particularly for sensitive technical topics.
• Not necessary for simple tasks: Fable's additional reasoning capability can be unnecessary for straightforward work such as short summaries, basic questions, or routine content generation. Using a smaller model can be more efficient for these tasks.
• Availability can vary: Access to Fable depends on the Claude plan, API setup, and platform you use. API users also pay separately for their token usage, so availability within a subscription should not be confused with unlimited API access.

How Does Claude Fable 5's Safeguard Fallback Work?

Because Mythos-class capabilities carry real misuse risk, Fable 5 ships with classifiers covering cybersecurity, biology/chemistry, and distillation. When a request trips one, it's automatically answered by Claude Opus 4.8 instead, and the user is told. Anthropic reports this happens in fewer than 5% of sessions; over 95% of Fable sessions involve no fallback at all.

This is also exactly why the starred benchmark rows above matter: on those cyber/bio tasks, the published high score is Mythos 5's, while Fable 5, with safeguards active, performs closer to Opus 4.8.

What you need to know as a developer:

• If your workload touches security research, bio/chem, or anything the classifiers read as distillation, expect a slice of responses to come from Opus 4.8.
• API customers must configure Anthropic's new Fallback API, on the API it isn't fully automatic the way it is in the Claude apps.
• You aren't charged Fable prices for rerouted requests.
• There's a 30-day data-retention requirement on all Mythos-class traffic, used only for safety monitoring (not training).

A gateway helps here: you can log which requests fell back, route them consistently, and keep the behavior uniform across your app.

Claude Fable 5 pricing

Claude Fable 5.1 is priced as a premium model, with a higher per-token cost than Anthropic's lower-priced models. The current standard pricing is $10 per million input tokens and $50 per million output tokens.

The standard input and output rates apply when Claude processes tokens normally. Prompt caching is priced separately. Cache writes cost more than standard input tokens, but subsequent cache reads are much cheaper at $0.25 per million tokens. This can make caching useful when you repeatedly send the same large prompts, system instructions, or context.

Fable 5.1 also supports Anthropic's Batch API, which provides a 50% discount on input and output token pricing for eligible asynchronous workloads.

The headline price does not necessarily represent the final cost of every task. A long-running agentic workflow can generate a large number of input and output tokens, while effective caching can reduce the cost of repeatedly processing the same context. Your actual spend therefore depends on the length of the prompts, the amount of output generated, how often context is reused, and whether you use standard or batch processing.

If you are comparing prices with older references to Claude Fable 5, keep in mind that Fable 5.1 is the current version. The pricing table should therefore use Fable 5.1 rather than presenting launch-period Fable 5 pricing as current.

Simplify Claude Fable 5.1 deployment with TrueFoundry

Manage model access, usage, budgets, and observability from one platform as you scale AI workloads.

Book a Demo

How to access the Claude Fable 5 API

Fable 5 is available today on:

• Claude API (Claude Platform), model string claude-fable-5
• Amazon Bedrock, Google Cloud Vertex AI, Microsoft Foundry
• TrueFoundry AI Gateway

Option A, Call the Claude API directly

Grab an API key from the Claude Console and call the model with the string claude-fable-5. Anthropic's quickstart walks through authentication and your first request: see the Claude API docs.

Calling the API directly is perfect for a quick test or a single app. The moment Fable 5 goes into production across more than one team, you need access control, budgets, and fallbacks, and that is where a gateway comes in.

Option B, Call Fable 5 through the TrueFoundry AI Gateway (recommended for production)

Direct API access is enough to prototype. Running a premium, token-hungry model like Fable 5 across an organization is a different problem: you have to control who can use it, cap spend, handle the safeguard fallbacks consistently, and keep one view of cost across providers. A gateway sits between your apps and the model and handles all of that.

How enterprises are actually using it, and why a gateway is better

• Cost governance on an expensive model. At $10/$50 per million tokens, an ungoverned Fable 5 rollout can run up a bill fast. Teams set per-team and per-application budgets, rate limits, and virtual keys so spend is capped and attributable before the first invoice, not after.
• Smart routing between Fable 5 and cheaper models. Most enterprises do not send everything to Fable 5. They route the hard, long-horizon jobs to Fable 5 and default routine traffic to Opus 4.8 or a smaller model, getting frontier quality where it matters without paying frontier prices everywhere.
• Automatic fallbacks and high availability. Launch-week demand for Fable 5 is expected to be high and hard to predict. Teams configure automatic fallback to Opus 4.8 or another provider on errors, rate limits, or capacity issues, so the app never goes dark.
• One integration across every provider. Fable 5 is available on Anthropic, Amazon Bedrock, Google Vertex AI, and Microsoft Foundry. Through the gateway it is one OpenAI-compatible endpoint, so switching providers or models is a one-line change with no rewrite and no lock-in.
• Observability and governance built in. Every call is logged with token usage, latency, and cost per request, per team, per model, plus guardrails and audit trails, which is what security and platform teams require before a new model is approved for production.

In practice that means a platform team can turn Fable 5 on for the whole company in minutes, with spend controls and fallbacks already in place, instead of each team wiring up the raw API and hoping the bill stays sane.

Step 1: Add Claude Fable 5 to your gateway

In the TrueFoundry gateway, open your connected Anthropic provider, search for claude-fable-5 on the Models Selection screen, enable it (its $10 / $50 pricing is shown inline), and use Access Control to decide which teams and virtual keys can call it.

Step 2: Call it with the auto-generated snippet

Open the Playground, pick claude-fable-5, and copy the ready-made Usage code snippet. TrueFoundry generates it for OpenAI, LangChain, Node.js, cURL, LlamaIndex, and more, in both streaming and non-streaming modes.

The win: switching between Fable 5, Opus 4.8, GPT-5.5, or any other model is a one-line change, and every call is governed, logged, and observable. Add the X-TFY-LOGGING-CONFIG and X-TFY-METADATA headers to tag and track spend per team or feature.

When should you use Claude Fable 5?

Claude Fable makes the most sense when you have a task that is difficult, time-consuming, or requires several steps to complete. It is built for work where a quick answer is not enough and the model needs to reason through a problem, use tools, and keep track of context along the way.

You could use Fable for a large software migration, a complex coding project, deep research, detailed document analysis, or an agentic workflow that runs through multiple stages. It can also be useful when you need to work with a large amount of information at once, such as a large codebase or a collection of documents.

But you don't need Fable for every task. If you are writing a short email, summarising a few paragraphs, answering a simple question, or processing a large volume of routine requests, a smaller and cheaper Claude model may be enough. Paying for a more capable model does not automatically make a simple task better.

The easiest way to think about it is to match the model to the task. Use Fable when its stronger reasoning and ability to handle longer workflows can save you time or improve the result. For simpler work, a less expensive model can usually do the job without the additional cost.

If you are choosing a model for a new project, also consider Claude Fable 5.1, which is the newer version of Fable 5. It is the version to evaluate for new long-running coding, research, and other complex workloads.

Ready to put Fable 5 into production with cost controls and fallbacks from day one? Try Claude Fable 5 on the TrueFoundry AI Gateway. Book a demo.