Claude Fable 5
Fable 5 shipped on 9 June 2026 and at $10 in and $50 out it is the dearest widely released model Anthropic sells. It is first on the Arena text leaderboard, and second in our coding table, behind a model that charges half as much.
- first on the Arena text board
- 1M token context window
- knowledge cutoff January 2026
Last checked: This is a reference page. It is re-checked against the vendor sources and updated when a new version ships. What changed
Why the dearest Anthropic model comes second in the coding table
On WebDev Arena, Fable 5 scores 1628 and Opus 5 scores 1692. Sixty-four points, in favour of the model charging half as much per output token. On the general text leaderboard the order flips: Fable 5 leads at 1506 and Opus 5 sits seventh at 1494. That gap is twelve points.
So the coding gap is more than five times the conversation gap, and it points the other way. The division of labour falls out of those two numbers: send code to Opus 5, send long-form text and conversation to Fable 5. Doing both with one model means overpaying on one of them.
Its knowledge is older than the cheaper model's
Fable 5 reads reliably through January 2026. Opus 5 reads through May 2026. The model that costs double is four months further behind, and a pricing table shows none of that. Working on a framework that moved this year, you find out in the first two or three answers rather than in a benchmark.
The likely reason is release order: Fable 5 arrived in June and Opus 5 in July. For anyone choosing today the reason changes nothing, and the result changes plenty.
The column where Fable 5 scores zero
Our coding category carries a score-per-dollar column that divides the leaderboard rating by the output price. Fable 5 lands on 32.6, the lowest of every candidate with a verified price, so it takes nothing from that column's twenty points of weight. DeepSeek V4 Flash lands on 5660 in the same column. The arithmetic is open on the category page and you can check it.
Do not read that as an argument for whatever is cheapest. Fable 5 takes full marks in the tooling-maturity column and ships on five routes, from Anthropic's own API to Claude Platform on AWS and Microsoft Foundry. If your company already has a contract on one of those clouds, that column will matter to you more than the price does.
Specifications
| API model ID | claude-fable-5 Source |
|---|---|
| Context window | 1 million tokens Source |
| Max output | 128k tokens Source |
| Input price | $10 per million tokens Source |
| Output price | $50 per million tokens Source |
| Knowledge cutoff | 2026-01 Source |
| Where it ships | Claude API, Amazon Bedrock, Claude Platform on AWS, Google Cloud, Microsoft Foundry Source |
Using it from Iran
This column carries a date and says how it was checked, because most listicles guess it. Where we have only read a vendor policy page, the note below says exactly that.
- Reachable
- blocked
- Payment
- no working route
- Free tier
- no
- Measured on
How we checked: Iran is not on Anthropic supported-countries list. Read from Anthropic own page rather than measured on the network.
What it is good at
- First on the Arena text leaderboard at 1506, ahead of Opus 5
- A one million token window with 128k tokens of output
- Ships on five routes: the Anthropic API, Bedrock, Claude Platform on AWS, Google Cloud and Microsoft Foundry
- Takes full marks in the tooling-maturity column of the coding table
Where it falls short
- It costs twice what Opus 5 costs and still sits below Opus 5 on the web-dev leaderboard.
- Its knowledge cutoff is January 2026, four months older than Opus 5.
- It finishes last in our score-per-dollar column and takes none of that column's twenty points.
- No official route from Iran: it is absent from Anthropic's supported-countries list and Iranian cards do not pay.
- Anthropic published no absolute benchmark figures for this model either, so part of our table stays empty for it.
Our take
If you want one model for everything and cost is not the first question, Fable 5 is a defensible pick. But a team that spends most of its time writing code pays double with Fable 5 to score lower at that exact job. In that case we take Opus 5 and keep Fable 5 for writing and conversation.
Questions people actually ask
Is Fable 5 better than Opus 5
It depends on the job, and both answers have a number behind them. On WebDev Arena Opus 5 scores 1692 against Fable 5's 1628, at half the price. On the general text leaderboard Fable 5 is first at 1506 and Opus 5 is seventh at 1494.
What does Fable 5 cost
Ten dollars per million input tokens and fifty per million output tokens. That is API pricing, and it is separate from the monthly Claude app subscription, which is billed on its own.
Why does the pricier model rank lower at code
Because a higher price does not have to buy a higher score. Anthropic calls Fable 5 its most capable widely released model and the general text board agrees, but Opus 5 arrived four months later and passed it on the web-dev board.
Can you reach Fable 5 from Iran
Not officially. Iran is not on Anthropic's supported-countries list, so there is no legitimate signup and no payment with an Iranian card. We do not sell accounts and we do not recommend circumvention.
Sources
- Anthropic model overview vendor source 12 August 2026