Home / Guides / Claude Fable pricing
Claude Fable pricing: the API rate and what each plan includes
Claude Fable 5.2, released on 9 November 2026, costs $8 per million input tokens and $40 per million output tokens on the Claude API, twice Opus 5.5 and 20% less than Fable 5.1, which stays available at $10 and $50 as a legacy model. On a subscription, Max ($100 a month) and Ultra ($200) include Fable up to half of the weekly usage limit. Pro ($20) pays for Fable with usage credits at the API rate, except in the launch week: until 16 November, Pro includes it too. Free has no Fable at all.
The 50% is a share of the weekly limit you already have, not extra room, and Fable spends that limit faster than other Claude models.
Checked 10 November 2026 against claude.com/pricing, Anthropic's API pricing and Claude Fable models on your plan · By Soren Starck
| Plan | Price | Claude Fable (5.2, 5.1 and 5) |
|---|---|---|
| Free | $0 | Not available |
| Pro | $20 a month ($17 billed annually) | Usage credits, at API rates (included to 16 Nov, launch offer) |
| Max | $100 a month | Included, up to 50% of the weekly limit |
| Ultra | $200 a month | Included, up to 50% of the weekly limit |
| Team, standard seat | $25 a seat a month ($20 annual) | Usage credits (included to 16 Nov, launch offer) |
| Team, premium seat | $125 a seat a month ($100 annual) | Included, up to 50% of the weekly limit |
| Enterprise, usage-based, and the API | Per token | Fable 5.2 at $8 input, $40 output per million tokens |
Claude Fable 5.2 API pricing
Per million tokens, from Anthropic's pricing page, with Fable 5.1 and Opus 5.5 beside it:
| Per million tokens | Fable 5.2 | Fable 5.1 (legacy) | Opus 5.5 |
|---|---|---|---|
| Input | $8 | $10 | $4 |
| Output | $40 | $50 | $20 |
| Cache write, 5 minutes | $10 | $12.50 | $5 |
| Cache write, 1 hour | $16 | $20 | $8 |
| Cache read | $0.20 | $0.25 | $0.20 |
| Batch input / output | $4 / $20 | $5 / $25 | $2 / $10 |
The whole 1M-token context window is billed at the same rates, with no long-context surcharge, and Fable 5.2 can write up to 128K tokens in one response (Fable 5.2 model page). The Batch API halves every rate. Asking for US-only inference multiplies all of them by 1.1. Fast mode, up to 2.5 times faster, now exists for Fable 5.2 as well as Opus 5.5, at twice the price: $16 and $80 per million tokens on Fable 5.2.
Fable 5.2 cuts every Fable 5.1 rate by 20%. Fable 5.1 had already cut the cache read from Fable 5's $1 to $0.25 (Anthropic's 5.1 launch post put that saving at about 25% for typical work), and Fable 5.2 takes it to $0.20, the same as Opus 5.5.
What one Claude Code session costs on Fable
Anthropic's Claude Code cost docs use a sample session of 1,200 input tokens, 5,300 output tokens, 940,000 cache reads and 50,000 cache writes. At list prices that session costs $0.91 on Fable 5.2 ($0.010 input, $0.212 output, $0.188 cache reads, $0.50 cache writes), against $1.14 on Fable 5.1, $0.55 on Opus 5.5 and $0.27 on Sonnet 5.5. So Fable 5.2 runs at about 1.7 times the cost of Opus in a coding agent, not the 2 times its headline rates suggest, because cache reads cost the same on both.
Fable 5.2's default effort is high, and higher effort means more output tokens per task. The rates of every other model and how usage credits work are in Claude pay-as-you-go pricing.
Is Claude Fable included in Claude Pro?
Not normally. Anthropic's help article says Fable 5.2, 5.1 and 5 are not part of Pro's usage limits, and the same goes for standard seats on Team and seat-based Enterprise plans. You can still pick Fable on Pro: each message is paid from usage credits, at the API rates above. The exception is the Fable 5.2 launch offer: from 9 November through 16 November 2026 (11:59 pm PT), Pro and Team standard seats that were paying before 9 November can spend up to 50% of the weekly limit on any Fable model at no extra cost, and get one Fable limit reset that expires the same day. The older promotion that put Fable 5 inside Pro's limits ended on July 19, 2026. Pro and Max are compared plan by plan in Claude Pro vs Max.
How much Fable Max and Ultra include
Max, Ultra and premium seats on Team and seat-based Enterprise plans can spend up to 50% of the weekly usage limit on Fable at no extra cost. It is a cap inside the weekly limit, not an addition to it: Fable and every other model draw from the same weekly limit, and Anthropic says Fable uses it faster than the others. When the Fable half is spent you can keep going on Fable with usage credits, or switch to Opus or Sonnet for the rest of the week. For the Fable 5.2 launch these plans also get one to three Fable limit resets, by how long you have paid (under six months one, six to twelve months two, more than twelve three), usable until 8 March 2027; a reset gives back the weekly usage spent on Fable and refills the 5-hour window. How the 5-hour and weekly limits work, and when each resets, is in Claude Code rate limits explained.
Fable on Team, Enterprise and the API
On Team, the seat type decides: a premium seat ($125 a month, or $100 billed annually) includes Fable within half its weekly limit, a standard seat ($25, or $20) pays with usage credits. Seat-based Enterprise works the same way by seat, and standard seats need the organization to have turned usage credits on. Usage-based Enterprise plans and the Claude API pay the standard rate, $8 and $40 per million tokens, under the model ID claude-fable-5-2 (Fable 5.1, claude-fable-5-1, stays at $10 and $50). Team seat prices are in Claude Team pricing.
Fable 5.2 or Opus 5.5?
Anthropic's model guide says to start with Opus 5.5 for most workloads and to use Fable 5.2 for demanding reasoning and long-horizon agent work. Both have the same 1M-token context, the same June 2026 knowledge cutoff and the same “moderate” latency in Anthropic's table. On a Max or Ultra plan the price shows up as limit, not dollars: Fable draws down the same weekly limit faster than Opus, and only half of that limit can go to Fable.
Watching your Fable limit
Claude's Settings > Usage shows the weekly limit for all models and the Fable share separately, but only when you open it. SessionWatcher keeps both in view: a Weekly Fable card sits next to the 5-hour session and the weekly limit, each with its reset time, in the Mac menu bar and the Windows taskbar, so you see the Fable half running out before usage credits start paying for it. Claude is in every SessionWatcher plan. Where else to check Claude Code usage is in how to check Claude Code usage.
SessionWatcherSee the Fable half of your weekly limit
before credits kick in.
Native app for the Mac menu bar and the Windows taskbar. Track Claude, Codex and 20 more tools' usage, costs, and rate limits in real time.
“Fast, simple, and does exactly what it should. Definitely worth it.”
@nicojerome on GitHub
macOS 14+. Free for 7 days, no credit card.
Frequently asked questions
How much does Claude Fable cost?
On the Claude API, Claude Fable 5.2, released on 9 November 2026, costs $8 per million input tokens and $40 per million output tokens, with cache reads at $0.20 and 5-minute cache writes at $10. Fable 5.1, now a legacy model, keeps its $10 and $50. On a plan, Max ($100 a month) and Ultra ($200) include Fable up to 50% of the weekly limit, and Pro ($20) reaches it through usage credits billed at the API rates, except during the launch week to 16 November. Read on Anthropic's pages on 10 November 2026.
Is Claude Fable included in Claude Pro?
Not normally. On Pro, and on standard seats of Team and seat-based Enterprise plans, Fable 5.2, 5.1 and 5 are not part of the plan's usage limits: they run on usage credits at standard API rates. The exception is Fable 5.2's launch offer: from 9 November through 16 November 2026 (11:59 pm PT), Pro and Team standard seats with paid billing before 9 November can use up to 50% of the weekly limit on Fable at no extra cost, plus one Fable limit reset that expires the same day.
How much Fable do Claude Max and Ultra include?
Up to 50% of the weekly usage limit, at no extra cost, on Max, Ultra and premium seats of Team and seat-based Enterprise plans. It is a share of the same weekly limit, not 50% more: Fable uses that limit faster than other Claude models. For the Fable 5.2 launch these plans also get one to three Fable limit resets, by how long you have paid (under six months one, six to twelve two, over twelve three), usable until 8 March 2027.
Is Claude Fable more expensive than Opus 5.5?
Yes, twice on input and output: Fable 5.2 is $8 and $40 per million tokens, Opus 5.5 $4 and $20. Cache reads cost the same ($0.20), so a Claude Code session that mostly rereads cached context costs about 1.7 times as much on Fable 5.2. Anthropic's model guide still suggests starting with Opus 5.5 for most work.
Is Fable 5.2 cheaper than Fable 5.1?
Yes, by 20% on every rate: input $8 against $10, output $40 against $50, cache reads $0.20 against $0.25, 5-minute cache writes $10 against $12.50. Fable 5.1 stays available at its old price as a legacy model. The sample Claude Code session below costs $0.91 on Fable 5.2 and $1.14 on 5.1.
Can I use Claude Fable in Claude Code?
Yes, on any paid plan: Fable 5.2 needs Claude Code version 2.1.321 or later (Fable 5.1 needs 2.1.255). It counts the same way as in the Claude apps: inside the Fable share of the weekly limit on Max and Ultra, on usage credits on Pro. Fable 5.2's default effort is high, which spends more tokens per task than medium.
Sources
All read on 10 November 2026:
- Claude pricing: plan prices, Fable on each plan, the API rate card, batch, US-only inference and fast mode.
- Anthropic API pricing: Fable 5.2 and Fable 5.1 rates, 1-hour cache writes, batch and fast mode prices, long context.
- Claude Fable 5.2 model page: release date, model ID, context, output limit, default effort.
- Claude Fable 5.2 launch offers: the launch week on Pro and Team standard seats, Fable limit resets on Max, Ultra and premium seats.
- Claude Fable models on your plan: the 50% weekly share, usage credits on Pro and standard seats, the Claude Code version.
- Models overview: when to choose Fable over Opus, Fable 5.1 under legacy models.
- Introducing Claude Fable 5.1 (September 2026): the 5.1 cache-read price cut and Anthropic's cost estimates.