Claude Fable 5.2 and Mythos 5.2: Anthropic cuts the price 20% and speeds output up to 30%
Direct answer
On November 9, 2026, Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, two versions of the same model for the hardest technical work. Fable 5.2 is generally available, with cybersecurity and biology safeguards. Mythos 5.2 is the same model with fewer safeguards, only for organizations verified under Anthropic's Cyber Verification and Life Sciences Verification programs. Both cost $8 per million input tokens and $40 per million output tokens, 20% less than the 5.1 generation's $10/$50, and generate outputs up to 30% faster. Anthropic still recommends Claude Opus 5.5 as the starting point for most coding and knowledge work: Fable is for hard, open-ended tasks.
One model, two doors
- Fable 5.2, the general door: available in Claude products, the Claude API as claude-fable-5-2, and on Amazon Web Services, Google Cloud and Microsoft Azure. It carries cybersecurity and biology safeguards, and Anthropic says those safeguards falsely block about 55% fewer requests than in 5.1, at the same strength.
- Mythos 5.2, the verified door: the same model with fewer safeguards, only for organizations that pass cyber or life-sciences verification. It also powers Claude Security for Enterprise customers.
- Price per task: the standard price fell 20%, but Anthropic charges for the work done. An optional fast mode costs $16/$80 per million tokens and generates up to 2.5 times faster. In the independent Artificial Analysis evaluation, Fable 5.2 at maximum effort reached 65 points on the index and about $26.54 per task, roughly double 5.1 in that configuration, because it generates many more tokens per task.
- Context and scope: 1-million-token context, 128,000-token maximum output and a June 2026 knowledge cutoff. On Pro and Team plans, Fable is included within the existing weekly limit through November 16, 2026, with a 50% cap.
The demonstration Anthropic puts on the table
Anthropic says a team of 1,000 Fable 5.2 agents formalized the proof of Fermat's Last Theorem in the Lean language in 30 hours. The caveat matters: outside mathematicians have not yet reviewed that work. It is a demonstration of agents coordinating on a long task; the mathematics community has not verified it yet. For your business, the practical read is different: the frontier is coordinating many agents on a long task without them drifting off course.
What to do with Fable 5.2 in your business
- Start with Opus 5.5 and move up only when the work demands it: Anthropic recommends exactly that. Fable costs twice as much per token as Opus 5.5 ($8/$40 versus $4/$20). Using the most powerful model for everything is paying extra for work the standard model already handles.
- Measure cost per finished task: the per-token price fell 20%, but a hard task at maximum effort can cost twice as much as before. Your buying metric is cost per result, with cache hits and retries included.
- Fast mode is for bottlenecks: up to 2.5 times faster at twice the price. It makes sense when waiting time slows your team down.
- Mythos is only for organizations in those fields: the version with fewer safeguards requires verification and is built for cyber defenders and life scientists. If a vendor offers you Mythos without that process, be suspicious.
Note: this article summarizes Anthropic's November 9, 2026 announcement and coverage in TechCrunch, The Next Web, VentureBeat, The Decoder and Artificial Analysis. Prices are Anthropic's published prices. Performance figures are vendor-reported and have not been independently reproduced; the Fermat proof formalization has not been reviewed by outside mathematicians.
Want AI that works for your business?
Message us on WhatsApp and let's talk about automation and AI agents for your business.
WhatsApp inicIARelated articles
- Claude Opus 5.5: the most powerful AI now costs 40% less
- ChatGPT vs Claude vs Gemini: which one suits your business in 2026
- Google opened Gemini 4 Argon with launch pricing: $2 per million tokens until Nov 30
Frequently asked questions
What is the difference between Fable 5.2 and Mythos 5.2?
They are the same model. Fable 5.2 is generally available with cybersecurity and biology safeguards. Mythos 5.2 has fewer safeguards and is only available to organizations verified under Anthropic's Cyber Verification and Life Sciences Verification programs.
How much does Fable 5.2 cost?
$8 per million input tokens and $40 per million output tokens, 20% less than Fable 5.1's $10/$50. An optional fast mode costs $16/$80 and generates up to 2.5 times faster.
Should I switch from Opus 5.5 to Fable 5.2?
Anthropic recommends starting with Opus 5.5 for most coding and knowledge work. Fable 5.2 is for hard, open-ended technical tasks, where the higher per-token cost is justified by the result.
Did Anthropic's AI prove Fermat's Last Theorem?
Anthropic says 1,000 Fable 5.2 agents formalized the proof in Lean in 30 hours. Outside mathematicians have not yet reviewed the work, so it is a demonstration of agent capability, pending external review.
