llmrelay
// claude

Claude Fable 5

The narrative long-form specialist.

Claude Fable 5 is the narrative and creative specialist in the Claude line. It is purpose-built for long-form writing, character consistency, and voice - the kind of work where Opus and Sonnet give you technically correct output that reads flat.

Anthropic positions it as the model for fiction, marketing copy, and any work where tone and flow matter as much as facts. It carries the same 1M-token context window as Opus 5 and Sonnet 5, so it can hold an entire manuscript or campaign brief in context while drafting.

We sell it at half Anthropic’s list price - $5.00 per million input tokens against $10.00. That puts it at the same per-token rate as Opus 5, which means the choice between them is about the kind of output you want rather than the budget.

Official list
llmrelay
Input / M tokens
$10.00
$5.00
Output / M tokens
$50.00
$25.00
Context window
—
1,000,000 tokens
Max output
—
128,000 tokens

What it costs you per month

Real-world budget scenarios. Numbers are simple sums — official list price vs llmrelay's 50% off tier.

Usage scenario
Official
llmrelay
Light coding (1M in / 200K out per month)
$20.00
$10.00save $10.00
Heavy Cursor / Cline user (50M in / 5M out per month)
$750.00
$375.00save $375.00
Production RAG (500M in / 20M out per month)
$6000.00
$3000.00save $3000.00

What it's good at

Best for

Pick Claude Fable 5 when

  • +Long-form narrative writing: fiction, scripts, newsletters, campaign content. Fable is tuned for voice and flow in a way the reasoning models are not.
  • +Character consistency across a long work. If you are drafting a novel or a campaign with recurring voices, Fable holds characterisation better than Opus.
  • +Marketing and brand copy where tone is load-bearing. Product descriptions, landing pages, email sequences - work where sounding human is part of the job.
  • +You need creative output at volume and Opus feels overpriced for the task. Fable costs the same per token but is built for this workload.

Choose something else when

  • !Coding, reasoning, or agent work. Fable is not a reasoning model. If you need it to debug code or plan a multi-step workflow, use Opus or Sonnet.
  • !Structured output or API responses. Fable’s tuning optimises for natural language, not JSON schema compliance or function calling.
  • !Work where factual accuracy is the only thing that matters. Technical documentation, legal copy, financial summaries - those are Opus jobs.
  • !High-volume batch work on a budget. At $5/M input Fable is expensive. If the task is classification or extraction rather than narrative, use Sonnet or Haiku.

Questions people ask about Claude Fable 5

How much does Claude Fable 5 cost?

Anthropic’s list price is $10.00 per million input tokens and $50.00 per million output tokens. On llmrelay it is $5.00 and $25.00 - exactly half list. That is the same input rate as Claude Opus 5, so the choice between them is about capability rather than cost.

Should I use Fable 5 or Opus 5 for writing?

Fable 5 if the output is narrative, marketing, or creative - anything where voice and flow matter. Opus 5 if it is technical, structured, or factual. The two models cost the same per token on llmrelay, so there is no budget reason to compromise.

Can Fable 5 write code?

It can, but it should not. Fable is tuned for natural language, not reasoning or structured output. Use Opus 5, Sonnet 5, or one of the GPT-5.6 models for coding work.

What is the context window on Claude Fable 5?

1,000,000 tokens input, with up to 128,000 output tokens. That is the same window as Opus 5 and Sonnet 5, so you can put an entire manuscript or campaign brief in context.

Try Claude Fable 5 at half the price

Free to create an account, no subscription, $10 minimum top-up — enough to run this model against a real task and judge quality yourself.

Get API key →

Compare Claude Fable 5 against the alternatives

The comparisons this model appears in, the models nearest it on price, and the full rate card.