Independent newsroom

Wednesday, September 2, 2026

The File

Every claim traces to a primary source. Uncertainty is labeled. Corrections are public.

AI · Wednesday, September 2, 2026

Anthropic releases Claude Fable 5.1 and Claude Mythos 5.1

The File desk · Sep 2, 2026, 2:20 PM UTC

Status

Confirmed from Anthropic’s Sept. 1, 2026, announcement, API docs, and system card. Performance numbers and “world’s most advanced” wording are Anthropic’s; they are not independently verified here.
  • Confirmed
    On Sept. 1, 2026, Anthropic published “Introducing Claude Fable 5.1 and Claude Mythos 5.1,” listed the same item in its newsroom, and posted Claude API docs and a system card dated Sept. 1. Docs say the general-release model ID is claude-fable-5-1, released September 1, 2026.
  • Unverified
    Independent replication of Anthropic’s benchmark table was not fetched. When the Cyber Verification Program will include Mythos-class models is stated only as “in the near future.”
On Sept. 1, 2026, Anthropic published “Introducing Claude Fable 5.1 and Claude Mythos 5.1.” It says the two are the same underlying model with different safeguards: Fable 5.1 is generally available; Mythos 5.1 is limited to trusted-access programs for cybersecurity and life sciences. API model ID for the general release is claude-fable-5-1. Docs say released September 1, 2026; 1M token context; 128K max output; $10 / $50 per million input/output tokens, same as Fable 5 except cache reads at $0.25 per million tokens. Anthropic says typical token-billed workloads cost an estimated 25% less than Fable 5 because of cheaper cache reads, and up to about 45% less for highly agentic work. Mythos 5.1 is invitation-only (Project Glasswing / Cyber Verification Program and Life Sciences Verification Program); Anthropic says it is currently limited to a set of U.S. organizations. Knowledge cutoff in the docs: June 2026. Company claims such as “world’s most advanced models for coding and knowledge work” are Anthropic’s; they are not treated here as independently verified. Benchmarks in Anthropic’s table (Terminal-Bench-Science 0.1 Fable 5.1 52.6% vs Fable 5 24.7%; Terminal-Bench 4.0 Fable 5.1 55.8% / Mythos 5.1 60.9%) are Anthropic’s, with the company’s footnotes on setup and safeguard interventions.
Anthropic wrote: “We’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They’re the world’s most advanced models for coding and knowledge work—and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress.” That “world’s most advanced” claim is Anthropic’s. This file does not treat it as independently verified.
The same Sept. 1, 2026, announcement says: “Claude Fable 5.1 and Claude Mythos 5.1 are the same model, but with different levels of safeguards. Fable 5.1 is generally available, while Mythos 5.1 is available only through our trusted access programs; its safeguards are specifically designed to support work in cybersecurity and the life sciences.” Anthropic’s newsroom lists the item under Announcements, Sept. 1, 2026.
Claude API docs for Fable 5.1 say: Latest. Released September 1, 2026. Model ID: claude-fable-5-1. Context window 1M tokens; max output 128K tokens; input $10 / MTok; output $50 / MTok; cache read $0.25 / MTok. Reliable knowledge cutoff: Jun 2026. Anthropic says Fable 5.1 is priced the same as Fable 5 except cache reads, and that typical token-billed workloads cost an estimated 25% less than Fable 5, with savings up to about 45% for highly agentic work.
Mythos 5.1 is invitation-only. The announcement names the Cyber Verification Program (CVP) and the Life Sciences Verification Program, and Project Glasswing in the docs. Anthropic says Mythos 5.1 “is currently only available to a set of US organizations.” It says the CVP currently covers certain Opus- and Sonnet-class models and “in the near future” will also include Mythos-class models. Enterprise Frontier Safeguards (EFS) “will be made available to enterprise customers in phases, beginning later this fall”; until then, Anthropic says eligible customers can use Fable 5.1 with zero data retention.
Anthropic’s system card, dated Sept. 1, 2026, states: “This system card describes Claude Fable 5.1 and Claude Mythos 5.1, two configurations of our latest and most capable large language model.” It says Fable includes extra dual-use safeguards in biology and cybersecurity and that Mythos relaxes some of those for vetted users. It also says Claude Security, available to Claude Enterprise customers, is powered by Mythos 5.1. This file does not copy exploit or weapons how-to from the system card.
Anthropic’s announcement table lists Terminal-Bench-Science 0.1 at 52.6% for Fable 5.1 versus 24.7% for Fable 5, and Terminal-Bench 4.0 at 55.8% for Fable 5.1 and 60.9% for Mythos 5.1. Footnotes and the system card attribute those scores to Anthropic’s setup (including Claude Code, effort, and, on some tasks, safeguard interventions). They are not independently replicated here.

What is still unknown or disputed

Primary sources

Every claim in this story is drawn from the documents below. If a fetch failed, that is recorded on the card.

  1. Source 1

    Introducing Claude Fable 5.1 and Claude Mythos 5.1

    Anthropic · September 1, 2026

    We’re introducing Claude Fable 5.1 and Claude Mythos 5.1. They’re the world’s most advanced models for coding and knowledge work—and their research capabilities offer an early glimpse of how AI models will contribute to scientific progress. Claude Fable 5.1 and Claude Mythos 5.1 are the same model, but with different levels of safeguards. Fable 5.1 is generally available, while Mythos 5.1 is available only through our trusted access programs. Currently, it is only available to a set of US organizations. Cache reads now cost 75% less, or $0.25 per million tokens. For typical workloads, costs are reduced by around 25% relative to Fable 5. For complex coding and highly agentic tasks, the savings could be up to around 45%. Terminal-Bench-Science 0.1: Fable 5.1 52.6%, Fable 5 24.7%. Terminal-Bench 4.0: Fable 5.1 55.8%, Mythos 5.1 60.9%.

    https://www.anthropic.com/claude-fable-and-mythos-5-1

  2. Source 2

    Newsroom

    Anthropic · September 1, 2026

    Introducing Claude Fable 5.1 and Claude Mythos 5.1. Announcements Sep 1, 2026. Our most advanced models for coding and knowledge work. Their research capabilities also offer an early glimpse of how AI models will contribute to scientific progress.

    https://www.anthropic.com/news

  3. Source 3

    Claude Fable 5.1 overview

    Anthropic (Claude API docs) · September 1, 2026

    Latest. Released September 1, 2026. Model ID: claude-fable-5-1. Context window: 1M tokens. Max output: 128K tokens. Input pricing: $10 / MTok. Output pricing: $50 / MTok. Cache read: $0.25 / MTok. Reliable knowledge cutoff: Jun 2026. Claude Mythos 5.1 offers the same capabilities to Project Glasswing participants only.

    https://platform.claude.com/docs/en/models/fable-5-1/overview

  4. Source 4

    What’s new in Claude Fable 5.1

    Anthropic (Claude API docs) · September 1, 2026

    Claude Fable 5.1 and Claude Mythos 5.1 are priced the same as Claude Fable 5, except for cache reads: Base input $10 / MTok; Cache reads $0.25 / MTok; Output $50 / MTok. Claude Mythos 5.1 is offered only to approved customers in Project Glasswing.

    https://platform.claude.com/docs/en/models/fable-5-1/whats-new-fable-5-1

  5. Source 5

    Claude Fable 5.1 & Claude Mythos 5.1 System Card

    Anthropic · September 1, 2026

    This system card describes Claude Fable 5.1 and Claude Mythos 5.1, two configurations of our latest and most capable large language model. Claude Fable 5.1 is available for general use, and includes additional safeguards that prevent it from performing certain tasks in high-risk, dual-use domains such as biology and cybersecurity. Claude Mythos 5.1 is the same model with more permissive safeguards in these domains. Direct access to Mythos 5.1 is limited to vetted individuals and organizations through our trusted access programs. Its capabilities also power Claude Security, available to all Claude Enterprise customers.

    https://www-cdn.anthropic.com/0339e6a7c5c7b87f5c07798616dc32c215d14235/Claude%20Fable%205.1%20&%20Claude%20Mythos%205.1%20System%20Card.pdf