Claude Fable 5.2 Launch: 20% Cheaper, Plus a Week of Fable on Claude Pro
Anthropic's most capable model just got faster and cheaper, and now undercuts GPT-6 Astra. Subscribers get a launch bonus too, but the Pro offer ends on 16 November.
11 November: Epoch AI has now scored Fable 5.2 on its Capabilities Index: 174, the highest of any model it tracks. The section on Fable or Opus is updated.
Short answer: Claude Fable 5.2 costs $8 per million input tokens and $40 per million output tokens on the Claude API. That's 20% less than Fable 5.1, and it's the regular price, not a promotion. Cached input costs $0.20 per million, and batch requests are half price. The time-limited deal is for Claude subscribers: until 16 November, Pro plans can use Fable within their normal weekly limits, and Max and Ultra subscribers get up to three free Fable limit resets.
Anthropic released Fable 5.2 on 9 November 2026, alongside Claude Mythos 5.2. Fable 5.1 cost the same as OpenAI's GPT-6 Astra; Fable 5.2 undercuts it by a fifth. Live rates for every Fable model are on our Claude Fable pricing page.
Claude Fable 5.2 price at a glance
| API model ID | claude-fable-5-2 |
|---|---|
| Input | $8 per 1M tokens |
| Output | $40 per 1M tokens |
| Cached input | $0.20 per 1M tokens (2.5% of the input price) |
| Cache writes | $10 per 1M tokens for 5 minutes, $16 for 1 hour |
| Batch | 50% off: $4 / $20 per 1M tokens |
| Fast mode | $16 / $80 per 1M tokens, for up to 2.5 times faster output |
| Context window | 1M tokens, at the same rate throughout |
| Maximum output | 128K tokens |
| Available on | The Claude API, Amazon Web Services, Google Cloud and Microsoft Azure |
From Anthropic's pricing and models pages, checked 10 November 2026.
What's new in Fable 5.2, besides the price
- Faster. Anthropic says it writes output up to 30% faster than Fable 5.1. The optional fast mode goes further, up to 2.5 times faster, at double the price.
- Fewer refusals. Fable keeps its cybersecurity and biology safeguards, but in Anthropic's early testing they blocked about 55% fewer requests than on Fable 5.1.
- Built for big agent jobs. Anthropic's showcase: 1,000 Fable 5.2 agents formalised the proof of Fermat's Last Theorem in 30 hours. At that scale, a 20% cut in the price per token is real money.
Claude Mythos 5.2 is the same model with fewer cybersecurity and biology safeguards. It costs the same, but only organisations accepted into Anthropic's Cyber or Life Sciences Verification Programs can use it.
Claude Fable 5.2 launch offers: what each plan gets
The offers cover every Fable model (5.2, 5.1 and 5) in the Claude apps and Claude Code. They don't apply to the API.
| Plan | What you get | Ends |
|---|---|---|
| Pro, and standard seats on Team | Fable included in your weekly limits for a week, up to 50% of them, plus one Fable limit reset | 16 Nov 2026, 11:59 pm PT |
| Max, Ultra, and premium seats on Team | One to three Fable limit resets, depending on how long you've subscribed | Resets expire 8 Mar 2027 |
From Anthropic's help article on the Fable 5.2 launch offers, read 10 November 2026. Premium seats on legacy seat-based Enterprise plans get the Max offer. The Free plan, other Enterprise plans and HIPAA-ready Enterprise plans aren't included.
- On Pro, this is the big one. Fable normally costs Pro users extra, through paid usage credits. Until 16 November it draws on your plan's weekly limits instead, at no extra cost. Pick Fable in the model picker (your default doesn't change), and keep an eye on Settings, then Usage: Fable uses up limits faster than other models.
- How many resets you get: one if you've paid for under six months, two for six to 12 months, three for longer. It counts from your first paid bill, even if you cancelled and came back.
- What a reset does: it gives back the weekly usage you've spent on Fable and, if your plan has one, refills the five-hour window. It doesn't raise your limits. Use it any time before it expires: Settings, then Usage, then "Reset for free".
- Who qualifies: accounts with paid billing in place before 9 November 2026 at 10:00 am PT.
Thinking of upgrading for Fable? Our Claude Code plans page sets Pro, Max and Ultra against the API cost of the same work.
Claude Fable 5.2 vs GPT-6 Astra and Opus 5.5: price
| Model | Input / output per 1M | Cached input |
|---|---|---|
| Claude Fable 5.2 | $8.00 / $40.00 | $0.20 |
| Claude Fable 5.1 | $10.00 / $50.00 | $0.25 |
| GPT-6 Astra | $10.00 / $50.00 | $1.00 |
| Claude Opus 5.5 | $4.00 / $20.00 | $0.20 |
Two numbers matter most. Against GPT-6 Astra, Fable 5.2 is 20% cheaper per token and five times cheaper on cached input. Against Claude Opus 5.5, it costs twice as much per token, but its cached input costs exactly the same.
That cache price is why agents feel the difference less than the headline suggests: most of what an agent sends at each step is a re-read of earlier steps. Here's a coding agent running 40-step tasks with 90% of repeated input cached:
| Model | Per task | Per task, no caching | Per month |
|---|---|---|---|
| Claude Opus 5.5 $4.00 / $20.00 per 1M | $1.94 | $9.96 | $853 |
| Claude Fable 5.2 $8.00 / $40.00 per 1M | $3.46 | $19.93 | $1,521 |
| Claude Fable 5.1 $10.00 / $50.00 per 1M | $4.32 | $24.91 | $1,901 |
| GPT-6 Astra $10.00 / $50.00 per 1M | $5.90 | $24.91 | $2,598 |
The agent calculator's coding agent example: 40 steps a task, 90% of repeated input cached where the model has a cache price, 20 tasks a day for 22 days. List prices checked 2–11 Nov 2026; this table updates automatically.
Side by side, with three workloads each: Claude Fable 5.2 vs GPT-6 Astra and Claude Fable 5.2 vs Opus 5.5.
Should you use Fable 5.2 or Opus 5.5?
Anthropic's own advice is to start with Claude Opus 5.5 for most coding and knowledge work, and move up to Fable only when a task needs it. Its launch results make a case for Fable on hard agentic work, though:
- Coding: on Anthropic's FrontierCode 2.0 test, Fable 5.2 scored 49.1% at its highest effort setting, against 45.5% for Opus 5.5 and 42.8% for GPT-6 Astra.
- Cost per task: Anthropic says that at High effort, Fable 5.2 scored about level with Opus 5.5's best result for roughly half the cost per task.
These are Anthropic's figures, but the independent check agrees on the order. On Epoch AI's Capabilities Index, Fable 5.2 scores 174 (range 170–179), the highest of any model Epoch tracks and clearly above Claude Opus 5.5 and GPT-6 Astra, both at 169. A higher score doesn't mean every task needs it: run both on a sample of your real work and compare the cost per finished task, not the price per token.
Upgrading from Fable 5.1
- Switching saves 20% on input, output and cached input. Change the model ID to
claude-fable-5-2. - There's no rush. Fable 5.1 stays available at $10 / $50, and Anthropic's deprecations page says it won't be retired before 1 September 2027.
- Blocked requests still get an answer. When the safeguards block a request, an Opus model answers it instead: Opus 4.8 for cybersecurity, Opus 5 for biology.
- Data retention: unless your organisation uses Anthropic's new Enterprise Frontier Safeguards, prompts and responses are kept for 30 days for safety monitoring.
Questions
How much does Claude Fable 5.2 cost?
$8 per million input tokens and $40 per million output tokens on the Claude API, with cached input at $0.20 and batch requests at half price. Fast mode costs $16 / $80.
Can I use Claude Fable 5.2 on the Pro plan?
Until 16 November 2026 at 11:59 pm PT, yes: Pro plans can use Fable models for up to 50% of their weekly limits at no extra cost, plus one Fable limit reset. After that, Fable on Pro runs on paid usage credits again. Max and Ultra include Fable as standard.
When does the Fable 5.2 launch offer end?
The Pro and standard Team week ends on 16 November 2026 at 11:59 pm PT, and its reset expires at the same time. Resets on Max, Ultra and premium Team seats last until 8 March 2027.
Is Claude Fable 5.2 cheaper than GPT-6 Astra?
Yes, per token: $8.00 / $40.00 against $10.00 / $50.00 per 1M input and output tokens, and $0.20 against $1.00 for cached input.
What's the difference between Claude Fable 5.2 and Mythos 5.2?
They're the same model at the same price. Fable 5.2 has stronger cybersecurity and biology safeguards and is open to everyone. Mythos 5.2 has fewer safeguards and is only for organisations in Anthropic's verification programs.
Sources: Anthropic's announcement, API pricing, models overview and model deprecations, and the Claude Help Center article on the Fable 5.2 launch offers. Read 10 November 2026.