Claude Fable 5.2: $8/$40 Price Cut and What Changed vs 5.1

Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on 9 November 2026. Fable 5.2 costs $8 per million input tokens and $40 per million output tokens, down from $10 and $50 for Fable 5.1, and Anthropic says output is up to 30% faster. Anthropic's own docs still tell most teams to start with Opus 5.5, not Fable.

Checked against Anthropic's pages on 13 November 2026. Benchmark figures are Anthropic's own and have not been independently reproduced.

Key Takeaways

• Fable 5.2 is real and confirmed: Anthropic's announcement, its pricing page and SD Times all describe the 9 November launch. The API ID is claude-fable-5-2.
• Standard price is $8 input and $40 output per million tokens, a 20% cut from Fable 5.1. Prompt-cache reads cost $0.20 per million, down from $0.25.
• Anthropic's benchmark table shows large gains over Fable 5.1 on agentic work, but only a narrow lead over Opus 5.5 on its knowledge-work measure.
• Fable and Mythos 5.2 are the same underlying model. Fable has stronger safeguards and is open to everyone; Mythos is limited to vetted organizations.
• Anthropic still recommends starting with Opus 5.5 at half the price, and reserving Fable for the hardest work.
• On Pro and standard Team seats, Fable is not part of plan limits except during a launch offer that ends 16 November 2026.

What was announced, and how I confirmed it

A search result had pointed to a "Claude Fable 5.2" on a single low-quality site, so I did not trust it. The claim holds up, though. Anthropic's news page lists "Introducing Claude Fable 5.2 and Claude Mythos 5.2" under 9 November 2026, the Claude Platform pricing and models pages list claude-fable-5-2 as active, and SD Times reported the launch the same day. A few aggregator pages that surfaced in my searches still treat 5.2 as a rumor. They look like stale snapshots, and I did not rely on them.

Anthropic's announcement says 5.2 improves coding, reasoning and vision, writes more clearly according to testers, and generates output up to 30% faster than 5.1. It adds an optional fast mode, up to 2.5 times faster, at $16 input and $80 output per million tokens. It is available on all platforms, including AWS, Google Cloud and Microsoft Azure.

Pricing next to the rest of the Claude line

All figures are per million tokens, from Anthropic's pricing and models pages, as of 13 November 2026.

Two details matter for budgeting. First, the Batch API takes 50% off input and output, but fast mode cannot be combined with batch. Second, every model on this table has adaptive thinking, and on Fable 5.2 and Opus 5.5 it is always on, so hidden reasoning tokens are billed as output. My cost estimate below leaves them out, which means real bills will be higher.

What the benchmarks show

Anthropic's launch table compares Fable 5.2, Fable 5.1 and Opus 5.5. These are vendor-reported numbers on benchmarks I have not run myself, so treat them as directional.

The pattern is worth reading slowly. Against Fable 5.1 the jumps are big. Against Opus 5.5 they are real but smaller, and on AA-Briefcase, a knowledge-work measure, the rubric scores are 66.1% for Fable 5.2 and 61.2% for Opus 5.5. On Anthropic's Elo view of that same benchmark, Opus 5.5 scores 1800 and Fable 5.2 scores 1785, so Opus is slightly ahead. For marketing and content tasks, which look more like knowledge work than like multi-hour engineering, that is the number I would weigh most.

Safeguards, and why Mythos is still out of reach

Anthropic says the cyber and biology safeguards are as robust as on 5.1 but produce about 55% fewer false-positive blocks (about 60% fewer in cybersecurity and 40% in biology). Fable blocks high-risk cyber requests and most research-biology requests, and reroutes them to Opus 4.8 (cyber) or Opus 5 (biology). It also ships what Anthropic calls its strongest anti-distillation protections so far.

Mythos 5.2 is the same model with more permissive safeguards, available to vetted organizations through the Cyber Verification Program and the Life Sciences Verification Program. Claude Enterprise customers can also use Claude Security, which runs Mythos 5.2. I covered why the split exists in Fable vs Mythos. Marketers writing normal content will rarely see a reroute. If you publish security or biotech material, expect some.

Cost math: one thousand content pieces

Assume 1,000 pieces, each with 8,000 input tokens (brief, style guide, source notes) and 4,000 visible output tokens. This is my illustrative workload, not a benchmark, and it excludes thinking tokens.

The arithmetic is input millions times input price plus output millions times output price. Fable 5.2 is exactly twice Opus 5.5 on this workload and four times Sonnet 5.5. Whether that premium pays off depends on how often the cheaper model's draft needs a rewrite. My earlier Sonnet 5.5 vs Opus 5.5 vs Fable 5.1 comparison covers the review-cost side of that question.

Who gets Fable 5.2 on which plan

Anthropic's support pages say Fable is not part of plan usage limits on Pro and standard Team seats, and runs on usage credits there. Max, Ultra and premium Team seats can spend up to 50% of their weekly limit on Fable. A launch offer includes Fable on Pro and standard Team seats, capped at 50% of the weekly limit, from 9 to 16 November 2026 (11:59 pm PT). The full plan breakdown is in my post on Claude plan and usage-limit changes.

What this means for marketers and founders

• Do not make Fable your default. Anthropic's guidance calls Opus 5.5 its "strongest daily driver" and suggests Fable for the most ambitious work. My read is that this is honest, and the knowledge-work scores support it.
• Test on 20 of your own tasks this week. Run the same prompts on Opus 5.5 at medium effort and Fable 5.2, score outputs blind, and compare cost per accepted output. The protocol is in how to evaluate a new model.
• If you run Fable 5.1 in production, there is no deadline. Anthropic lists 5.1 as active with a retirement date of no sooner than 1 September 2027. Move when your test says so, and expect behavior differences because it is a new model ID.
• Restructure prompts for the cache. At $0.20 per million, a repeated brand bible is cheap to reuse. See prompt caching for marketing teams.
• Leave fast mode alone for batch work. It doubles the price and cannot be batched. It suits live drafting sessions, not overnight jobs.
• Enterprise admins: check your default. Per Anthropic's support page, Fable 5.2 stays off until an admin enables it, and seven days after launch it is added to roles that have any Fable model enabled, with defaults moving to 5.2 unless an admin opts out first.

What we do not know yet

• Independent results. Outside testers named in the announcement include the UK AI Security Institute and METR, but the benchmark table is Anthropic's.
• Real-world token use. Anthropic says output is up to 30% faster, but not how many total tokens a task consumes at each effort level.
• Writing quality. None of the benchmarks measures brand voice, so only your own samples can answer that.
• Whether the lower price lasts. Anthropic's Sonnet 5 pricing history shows rates can change, as I explain in my post on the cancelled Sonnet 5 price rise.

FAQ

Is Claude Fable 5.2 real?

Yes. Anthropic announced Claude Fable 5.2 and Claude Mythos 5.2 on 9 November 2026, and its pricing and models pages list claude-fable-5-2 as an active model. SD Times also reported the launch. A single low-quality site was the only source in my first search, which is why I checked Anthropic's own pages.

How much does Claude Fable 5.2 cost?

The standard API price is $8 per million input tokens and $40 per million output tokens, as of 13 November 2026. Fast mode is $16 and $80. Prompt-cache reads cost $0.20 per million.

Is Fable 5.2 cheaper than Fable 5.1?

Yes. Fable 5.1 costs $10 input and $50 output per million tokens, so 5.2 is 20% cheaper at standard rates. Cache reads fall from $0.25 to $0.20 per million.

Should I use Fable 5.2 or Opus 5.5?

Anthropic recommends starting with Opus 5.5 for most coding and knowledge work, and moving to Fable 5.2 when your tests on Opus 5.5 at higher effort still fall short. Opus 5.5 costs half as much. On Anthropic's own knowledge-work Elo score, Opus 5.5 edges Fable 5.2.

What is the difference between Fable 5.2 and Mythos 5.2?

They are the same underlying model. Fable 5.2 has stronger cyber and biology safeguards and is available to all users. Mythos 5.2 has more permissive safeguards and is limited to organizations vetted through Anthropic's verification programs.

Can I use Fable 5.2 on a Claude Pro plan?

Fable is not part of Pro usage limits in normal weeks and runs on usage credits. From 9 to 16 November 2026, a launch offer includes Fable on Pro, capped at 50% of the weekly limit. After that, it goes back to usage credits.

Does Fable 5.2 replace Fable 5.1?

Not automatically. Fable 5.1 remains active, with a retirement date of no sooner than 1 September 2027 according to Anthropic's deprecations page. You choose when to switch by changing the model ID.

Want help with this?

I'm Younus Fardeen, an organic growth, SEO and AEO strategist with 4+ years of marketing experience across edtech and startup teams. If you want help choosing and testing AI models for your content and growth workflows, see my work and reach me through the contact form at younusfardeen.com.