Back to home

What is Fable 5? Anthropic's Most Powerful AI Model Explained

Fable 5 (claude-fable-5) is Anthropic's newest flagship AI model — a tier above Claude Opus. Specs, capabilities, pricing, and how to try it.

Last updated: 2026-06-10

Fable 5 is Anthropic's newest and most powerful AI model — a brand-new intelligence tier that sits above the Claude Opus family. Its API model ID is claude-fable-5.

If Claude models used to come in three sizes — Haiku (fast), Sonnet (balanced), Opus (most capable) — Fable 5 introduces a fourth tier at the very top of the lineup.

Key specifications

| Spec | Fable 5 | | --- | --- | | Model ID | claude-fable-5 | | Context window | 1,000,000 tokens | | Max output | 128,000 tokens | | Input price | $10 per million tokens | | Output price | $50 per million tokens | | Thinking | Adaptive (automatic) | | Vision | Yes, high resolution |

What makes Fable 5 different

A new tier above Opus

Claude Opus 4.8 was previously Anthropic's most capable model. Fable 5 raises the intelligence ceiling further on reasoning, coding, knowledge work, and long-horizon agentic tasks — at 2× the price of Opus 4.8 ($10/$50 vs $5/$25 per million tokens).

1M-token context at standard pricing

Fable 5 accepts up to one million tokens of context — roughly 700,000 English words, an entire large codebase, or several books — in a single conversation, with no long-context price premium.

Adaptive thinking only

Fable 5 decides for itself when and how deeply to reason. There are no manual "thinking budgets" — you enable thinking: {type: "adaptive"} and the model scales its reasoning to the difficulty of the task. Notably, the older sampling parameters (temperature, top_p, top_k) are removed entirely, and explicitly disabling thinking is not supported — a deliberate simplification of the API surface.

Built for agents

Fable 5 is state-of-the-art on long-horizon autonomous work: multi-hour coding sessions, complex refactors, deep research, and agentic loops that finish without human correction. It pairs with effort controls (lowmax) and task budgets so you can tune depth versus cost.

How can you use Fable 5?

  1. Claude apps & Claude Code — Fable 5 powers Anthropic's own products for Pro/Max-tier users.
  2. Anthropic API — call model ID claude-fable-5 directly. See our Fable 5 API guide.
  3. Right here — our online playground lets you chat with Fable 5 in your browser, no API key required. New accounts get free credits.

How does it compare?

For a detailed head-to-head with the previous flagship, read Fable 5 vs Claude Opus 4.8. The short version: same API surface, higher ceiling, double the price — worth it when the task demands maximum intelligence.


fable-5.net is an independent resource and is not affiliated with Anthropic.