Model components / Fine-tuned / chat models

Claude Fable 5

Anthropic's first Mythos-class model for general availability (June 2026), sitting above Opus in capability. Same underlying weights as Claude Mythos 5 but wrapped with safety classifiers that route sensitive cybersecurity, biology/chemistry, and distillation queries to Claude Opus 4.8 (~5% of sessions). Built for long-horizon agentic coding, knowledge work, vision, and scientific research with a 1M-token context window and up to 128K output. Tops Cognition FrontierCode and Hebbia Finance among frontier models; state-of-the-art on CursorBench and FrontierBench per early partner testing. $10/$50 per Mtok via API, Bedrock, Vertex, and Foundry.

Anthropic Claude Fable 5 (model id claude-fable-5), released Jun 9 2026. Mythos-class tier above Opus; generally available counterpart to restricted Claude Mythos 5 (Project Glasswing). 30-day mandatory data retention (Covered Model). Verified via Anthropic Fable 5 launch post and API docs June 2026.

Openness

1 high confidence

1.0

weights: closed
data: closed
code: closed
license: Proprietary(API-only)

No weights, data or code released. Served through the Anthropic API only, so 1/closed on every dimension. Recorded here rather than in base_pretrained because a closed model has no observable base checkpoint to score, which is the convention #114 settled.

https://www.anthropic.com/news/claude-fable-5-mythos-5 recorded 2026-06-11
Mythos-class flagship, proprietary, claude-fable-5 via API and cloud marketplaces only
https://platform.claude.com/docs/en/about-claude/models/introducing-claude-fable-5-and-claude-mythos-5 recorded 2026-06-11
API model id claude-fable-5, no weights release, Covered Model retention policy

Adoption

4 low confidence

4.0

Directional market-position estimate (not a measured usage figure). Anthropic's most capable GA model, with day-one enterprise testimonials (Stripe, GitHub, Cursor, Cognition) and a staged subscription rollout. Capped at 4 rather than 5 because it launched Jun 9 2026 (~2 weeks before scoring), so sustained reach is unproven. Overrides the prior frozen-anchor convention (held null) for dominant closed incumbents.

https://www.anthropic.com/news/claude-fable-5-mythos-5 recorded 2026-06-11
launch testimonials and staged availability, no quantitative usage figures
https://openrouter.ai/rankings recorded 2026-06-23
public per-model token-share leaderboard where Claude flagships rank among the most-used models

Capability

5 high confidence

5.0

Anthropic-reported: highest among frontier models on Cognition FrontierCode (even at medium effort); top on Hebbia Finance Benchmark; SOTA on CursorBench and FrontierBench per partner testing; leads Opus 4.8 on long-horizon coding, knowledge work, vision, and memory tasks. Vendor-reported; underlying Mythos weights with classifier fallback on ~5% of sessions.

https://www.anthropic.com/news/claude-fable-5-mythos-5 recorded 2026-06-11
FrontierCode highest among frontier models; SOTA across software engineering, knowledge work, vision
https://www.anthropic.com/claude/fable recorded 2026-06-11
most capable generally available Claude model; long-horizon agentic flagship positioning

Unchanged since 2026-07-30 (last edited, not re-checked)