Kuponaut

Kuponaut › Blog

Claude Fable 5.2 pricing: the API price cut and the Fable week for Pro (November 2026)

Anthropic released Claude Fable 5.2 on 9 November 2026 and lowered its API price: $8 per million input tokens and $40 per million output tokens, against $10 and $50 for Fable 5.1. For one week, Pro plans and standard seats on Team plans can also use Fable models inside their plan, up to 50% of their weekly usage limits, until 16 November 2026 at 11:59 pm PT. This page lists the prices and the offer terms as Anthropic's own pages showed them on 11 November 2026.

By Fırat Papatya · Published · 6 min read

No affiliate links: Kuponaut has no partner link or code with Anthropic and earns nothing from this article. Anthropic's announcement, its launch-offers help article, its API pricing page and its plans page were read on 11 November 2026; prices are in USD before tax and can change. We have not benchmarked the model ourselves. This article was drafted with Claude, the product it describes, and checked by a person.

What Claude Fable 5.2 costs on the API

Anthropic's API pricing page lists Fable 5.2 at $8 per million input tokens and $40 per million output tokens. Fable 5.1 is still listed at $10 and $50, so the new model is 20% cheaper on both sides (our arithmetic from the two prices). The announcement also says Fable 5.2 generates output up to 30% faster than Fable 5.1.

The model ID is claude-fable-5-2. The pricing page says the full 1M-token context window is included at the standard price, so a long prompt does not move you to a higher rate. Claude Mythos 5.2, the same underlying model for approved organizations, is listed at the same prices.

The announcement says Fable 5.2 is available on Amazon Web Services, Google Cloud and Microsoft Azure as well. Amazon Bedrock and the Google-managed service set their own prices, so check their pages if you buy there.

Per million tokensFable 5.2Fable 5.1
Input$8$10
Output$40$50
Cache write, 5 minutes$10$12.50
Cache write, 1 hour$16$20
Cache read$0.20$0.25
Batch API input$4$5
Batch API output$20$25
Fast mode input / output$16 / $80not listed

USD. Source: Anthropic's API pricing page, read 11 November 2026.

Batch, caching and fast mode: what moves the bill

Batch API: half price. Requests that can wait are processed asynchronously with a 50% discount on input and output, which makes Fable 5.2 $4 and $20 per million tokens.

Prompt caching: a cache read on Fable 5.2 costs $0.20 per million tokens, 2.5% of the input price. Writing to the 5-minute cache costs $10 per million tokens, and Anthropic's page says that pays for itself after one cache read. If you send the same long system prompt or document again and again, this is the deepest discount on the page: $0.20 instead of $8. The page also says the Batch discount and caching can be combined.

Fast mode: double price. The announcement says fast mode speeds up token generation by up to 2.5 times and costs $16 per million input tokens and $80 per million output tokens. The pricing page says it cannot be used with the Batch API. Use it only when the waiting time costs you more than the tokens.

The launch offer: one week of Fable for Pro and Team standard seats

Normally the plans page lists Fable on the Pro plan as paid with usage credits, and as included up to half of the weekly limit on Max and Ultra. The Free plan has no Fable access.

For the launch, Anthropic's help article says all Fable models (Fable 5.2, 5.1 and 5) are included in the weekly usage limits of Pro plans and standard seats on Team plans from 9 November 2026 through 16 November 2026 at 11:59 pm PT, which is 07:59 UTC on 17 November. During that week you can spend up to 50% of your weekly usage limits on Fable models at no extra cost. After you use that 50%, or once the week ends, Fable usage is billed to usage credits if you have them turned on.

The same plans also get one Fable limit reset, which expires at the same moment. The resets go only to accounts that had paid billing in place before 9 November 2026 at 10:00 am PT. A subscription that started before then but is still in a free trial gets its reset once it converts to paid.

Your default model does not change. To use the offer, pick a Fable model from the model picker. The help article warns that Fable models use up your limits faster than other Claude models, so watch Settings > Usage during the week.

Plan or seatWhat you getEnds
Pro; standard seats on TeamAll Fable models included for one week, up to 50% of weekly usage limits, plus one Fable limit reset16 Nov 2026, 11:59 pm PT
Max; Ultra; premium seats on Team; premium seats on legacy seat-based EnterpriseOne to three Fable limit resets, by subscription lengthResets expire 8 Mar 2027, 11:59 pm PT
Free; standard seats on legacy seat-based Enterprise; self-serve Enterprise; usage-based EnterpriseNot includedNot applicable

Source: Claude Help Center, Claude Fable 5.2 launch offers, read 11 November 2026.

How a Fable limit reset works

A reset does not raise your limits. On the weekly limit it gives back the usage you have spent on Fable models and leaves usage on other models where it is. If your account has a five-hour limit, the reset refills that window completely. Anthropic's example: you have used 80% of your weekly limit, 50% on Fable and 30% on other models; after a reset you are back at 30%.

On Max and Ultra plans and on premium seats, the number of resets depends on how long you have been a paying subscriber: one for less than six months, two for six to 12 months, three for more than 12 months. The time is counted from your first paid bill, and a gap in the subscription does not restart it. On Team and Enterprise plans it is counted from the organization's first paid bill.

To use one, open Settings > Usage and select “Reset for free” under Limit resets. You can do this at any time before the reset expires, not only when you hit a limit. It cannot be undone, and a reset you have not used is lost if you cancel or downgrade.

Fable 5.2 or Opus 5.5: which one to pay for

Claude Opus 5.5 costs $4 per million input tokens and $20 per million output tokens, exactly half of Fable 5.2. Anthropic's own advice in the announcement is to start with Opus 5.5 for most coding and knowledge tasks and to move up to Fable only for the most ambitious, open-ended technical work.

On the cost side Anthropic gives one comparison. On its FrontierCode 2.0 coding benchmark, Fable 5.2 at High effort scored 45.3%, about level with the best Opus 5.5 score of 45.5%, at roughly half the cost per task. At the Xhigh effort level it reached 49.1%. These are Anthropic's measurements; we have not reproduced them.

Our own same-prompt test on the ChatGPT vs Claude vs Gemini vs Grok page was run on Fable 5.1. We have not rerun it on Fable 5.2 yet.

ModelInputOutputBatch inputBatch output
Claude Fable 5.2$8$40$4$20
Claude Opus 5.5$4$20$2$10
Claude Sonnet 5.5$2$10$1$5
Claude Haiku 5.5 (prompts up to 100,000 tokens)$0.10$0.50$0.05$0.25
Claude Haiku 5.5 (prompts over 100,000 tokens)$0.50$2.50$0.25$1.25

USD per million tokens. Source: Anthropic's API pricing page, read 11 November 2026.

What the Claude plans cost

These are the individual plan prices on Anthropic's plans page, opened from Turkey on 11 November 2026. The Fable column shows the normal state, outside the launch week.

PlanPriceFable access
Free$0None
Pro$20 a month, or $17 a month on the annual plan ($200 billed up front)Paid with usage credits; included up to 50% of weekly limits until 16 Nov 2026
Max$100 a month, 5x Pro usageIncluded, up to 50% of weekly limits
Ultra$200 a month, 10x Pro weekly usageIncluded, up to 50% of weekly limits

USD. Source: Claude plans and pricing page, read 11 November 2026.

Seven ways to spend less on Fable

1. Start on Opus 5.5 and switch to Fable only for the hard part of the job. It is half the price per token and it is what Anthropic recommends as the daily model.

2. Leave the effort level at Medium unless a task needs more. The announcement says Medium is the default in the Claude apps and in Claude Code, and that at Medium Fable 5.2 can finish challenging tasks with fewer tokens than Fable 5.1.

3. On the API, send work that can wait through the Batch API. That is 50% off.

4. Cache what you repeat. A long system prompt or a codebase summary that is read from cache costs $0.20 per million tokens instead of $8.

5. Keep fast mode off by default. It doubles the price.

6. On a subscription, check Settings > Usage before a long session. Fable uses your weekly limit faster than the other models, and a reset only helps if you use it before it expires.

7. If you are on Pro, decide before 16 November whether you want Fable to keep running on usage credits after the week ends, and check whether usage credits are turned on in your account.

Frequently asked questions

How much does Claude Fable 5.2 cost?

On the API it costs $8 per million input tokens and $40 per million output tokens. The Batch API halves that to $4 and $20, and fast mode doubles it to $16 and $80. Prices were read on Anthropic's pricing page on 11 November 2026.

Is Claude Fable 5.2 free?

Not on the Free plan. Pro plans and standard seats on Team plans have Fable included until 16 November 2026 at 11:59 pm PT, up to 50% of their weekly usage limits. After that Fable runs on usage credits on those plans. Max and Ultra include Fable up to 50% of weekly limits.

When does the Fable week for Pro end?

On 16 November 2026 at 11:59 pm PT, which is 07:59 UTC on 17 November. The one Fable limit reset that Pro and Team standard seats receive expires at the same time.

Is Fable 5.2 cheaper than Fable 5.1?

Yes. Fable 5.1 is listed at $10 input and $50 output per million tokens, Fable 5.2 at $8 and $40. That is 20% less.

How does the price compare with Claude Opus 5.5?

Opus 5.5 costs $4 input and $20 output per million tokens, half the price of Fable 5.2. Anthropic recommends starting with Opus 5.5 for most coding and knowledge work.

Is there a Claude promo code for Fable 5.2?

Kuponaut has no code and no partner link for Claude. The launch offer does not use a code: on an eligible plan you pick a Fable model from the model picker.

Links

A code-free copy of this article is also published on: Dev.to (canonical URL is this page).

Listed on