Claude Fable 5
Previous Fable. Same $10/$50 list, still served. Use 5.1 for new work.
Claude Fable 5 is the previous Fable snapshot. It still runs on llmrelay at $5.00 / $25.00 — half Anthropic list, same invoice as Fable 5.1. For new work, use Fable 5.1. Anthropic’s current Fable page is coding and knowledge work, not a writing sideline; 5.1 is the escalation above Opus 5.
This page stays up because pinned ids matter. If you already validated output against claude-fable-5, keep calling it. If you are starting a Claude Code job, a long-horizon review, or anything you would have called “the Fable” last month, the current id is claude-fable-5-1 (hyphen).
Same 1M context window and 128k max output as 5.1 and Opus 5. Anthropic still rates Fable slower than Opus, and adaptive thinking is always on, so easy prompts still pay for reasoning tokens. That is why Opus 5 at $2.50 / $12.50 remains the default on this key.
Claude Fable 5 still runs on llmrelay at the rates below, and we keep selling it — pinned versions matter when you have output you do not want to re-validate. For new work, Claude Fable 5.1 is the current pick at $5/M input.
What it costs you per month
Real-world budget scenarios. Numbers are simple sums — official list price vs llmrelay's 50% off tier.
What it's good at
- +Pinned Fable 5 snapshot
- +Same $5/$25 as 5.1
- +Still callable on channel 3
Best for
- — Reproducing Fable 5 output
- — Pinned agent runs
- — Not new Claude Code work
Pick Claude Fable 5 when
- +You already validated a pipeline against claude-fable-5 and do not want to re-check 5.1.
- +A contract or eval set names this exact id. Switching snapshots is a new test, not a free upgrade.
- +You need the previous Fable on the same prepaid key as Opus 5 and GPT-6 Astra — one model-id change, no new vendor.
Choose something else when
- !New work. Fable 5.1 is the current Fable at the same $5/$25, with Anthropic’s cheaper cache hits on their card and a June 2026 cutoff.
- !Everyday coding. Opus 5 is $2.50/M input. Anthropic tells you to start there and reach for Fable only when Opus evals fall short.
- !You typed this URL looking for a writing model. That was old copy on this page. Fable 5 was never a fiction-only SKU on the API; 5.1 makes the coding positioning explicit.
- !High-volume loops. At $5/M input plus always-on thinking, the bill grows faster than the rate card.
Questions people ask about Claude Fable 5
How much does Claude Fable 5 cost?
Anthropic list 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, the same as Fable 5.1. Opus 5 is $2.50 / $12.50, so Fable is the expensive Claude, not the writing tier.
Should I use Fable 5 or Fable 5.1?
Fable 5.1 for new work. Use Fable 5 only to reproduce output you already validated against this id. Both cost $5.00 / $25.00 here. The current id is claude-fable-5-1 (hyphen, not a dot).
Should I use Fable 5 or Opus 5?
Default to Opus 5 at $2.50 / $12.50. Reach for a Fable id when Opus 5 has already failed the specific task. If you are escalating, escalate to Fable 5.1, not this snapshot.
Is Fable 5 a writing model?
No. An earlier version of this page called it a narrative specialist. Anthropic sells Fable as the top Claude for hard coding and knowledge work, at twice the Opus rate. For new writing or coding, use Fable 5.1 or start on Opus 5.
What is the context window on Claude Fable 5?
1,000,000 tokens input, with up to 128,000 output tokens. Same window as Fable 5.1, Opus 5 and Sonnet 5.
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.
- Fable 5 vs Opus 5 — is the flagship worth 2×?Fable 5 is the previous Fable snapshot, still sold at $5/$25 — twice Opus 5.
- Claude Opus 5 pricing and specs$2.50/M in, $12.50/M out — half list. Current flagship Opus. A generation past 4.8, same list price.
- Claude Fable 5.1 pricing and specs$5.00/M in, $25.00/M out — half list. Anthropic’s current top Claude. Same $10/$50 list, cheaper cache reads.
- GPT-6 Astra pricing and specs$5.00/M in, $25.00/M out — half list. OpenAI’s current flagship. API, ChatGPT paid plans, half list here.
- Full price listEvery model we serve, at 50% of official list. No subscription, no volume gate.
- Cost calculatorPut your own monthly token volume in and see the bill against vendor list price.