# Claude Fable 5

> Claude Fable 5 is the previous flagship of Anthropic's Fable line — a Mythos-class model built for autonomous coding, sustained knowledge work, and agents that need to stay coherent over a long horizon. It carries a 1M-token context window with native prompt caching and vision input.

Canonical: https://routemux.com/anthropic/claude-fable-5
Provider: Anthropic

## Key facts

| Field | Value |
| --- | --- |
| Model ID | anthropic/claude-fable-5 |
| Official model ID | claude-fable-5 |
| Context window | 1M |
| Max output | 8K |
| Input modalities | text, image |
| Output modalities | text |
| Capabilities | Vision, Reasoning, Tool Calling, Coding, Cache, Strip Sampling Params |
| Upstream release | 2026-06-09 |
| Availability | available |

## Pricing

- RouteMux input: $2.8 per 1M tokens
- RouteMux output: $14 per 1M tokens
- Official input: $10 per 1M tokens
- Official output: $50 per 1M tokens
- Cache read: $0.28 per 1M tokens
- Cache write: $3.5 per 1M tokens
- Saving vs official: 72%
- Official price source: anthropic_x0.28, verified 2026-08-16

Billing is prepaid wallet, charged per successful request only.

## API access

Base URL: https://api.routemux.com

Supported protocols:

- anthropic_messages

## What is Claude Fable 5

Claude Fable 5 is the previous flagship of Anthropic's Fable line — a Mythos-class model built for autonomous coding, sustained knowledge work, and agents that need to stay coherent over a long horizon. It carries a 1M-token context window with native prompt caching and vision input.

Through RouteMux it is $2.80 per 1M input tokens against Anthropic's $10, and $14 against $50 on output. Same model, same Anthropic Messages API, billed from a prepaid wallet instead of an Anthropic account, with failed and timed-out requests not charged.

It has been succeeded by Claude Fable 5.1, which costs the same per input and output token but reads from cache at $0.07 per 1M tokens against $0.28 here. New projects should start on 5.1; this page stays for callers already running on Fable 5.

Fable sits above Opus 5 in the lineup and costs twice as much per token, so the honest way to choose between them is to try Opus 5 first. Fable earns its price on tasks where a partially-correct answer is worthless — long refactors that must not leave stubs, analysis where a missed detail invalidates the conclusion.

## Key capabilities

- **Previous Fable flagship** — Anthropic's flagship tier for autonomous coding and long-horizon work, now succeeded by Fable 5.1. This tier remains in service.
- **1M-token context with caching** — Native prompt caching, which matters more here than on cheaper models because the base rate is what caching discounts.
- **Vision input** — Images go in the same message as the prompt — screenshots, diagrams, scanned documents.
- **Anthropic Messages native** — No translation layer; the request is the same one you would send to Anthropic directly.

## Other variants in this family

- Claude Fable 5.1 (`anthropic/claude-fable-5-1`)

## Sources

- [Anthropic — Models overview](https://docs.anthropic.com/en/docs/about-claude/models)
- [Anthropic — Pricing](https://www.anthropic.com/pricing)

_Content reviewed by Jerry Fan on 2026-09-02._