Compare Start here

AnthropicLanguage modelSuperseded

Claude Fable 5

The first publicly available Mythos-class model, shipped behind new safety classifiers.

Compare head-to-head Race it in 3D

Overall Intelligence

#7

of 56
2.9 pts behind #1

0

races led

408

weeks at #1

85.8

score

Released: Jun 9, 2026; 4 mo ago
Context: 1M; 128K max output
Price: $10 / $50; per 1M in/out · $20.00 blended
Speed: 48 tok/s; 10s to first token
Modalities: text · image; outputs text
Weights: Closed; API and apps

The 30-second version

The first 'Mythos-class' tier above Opus. Fable 5 briefly led the overall race and set the bar for long-horizon autonomy.

Key takeaways

1Created a new top tier above Opus.
2Pushed autonomous task length sharply upward.
3Replaced by Fable 5.1 in August 2026.

Where it ranks

Current snapshot · illustrative scores · click any race to watch it

Fable 5Best on each axis

Overall

7Overall+2.9 pts

Coding

4Coding+3.1 pts 9Frontend & UI+35 Elo 3Backend & APIs+2.1 pts 4Distributed Systems+2.5 pts 4Full-Stack Apps+3.8 pts 4Debugging & Repair+3.7 pts 5Agentic Coding+3.7 pts 6Code Review & Security+3.9 pts 5Data Science & SQL+1.9 pts 5Mobile Apps+3.0 pts 3Terminal & DevOps+4.9 pts 6Competitive Programming+58

Hard Reasoning

5Hard Reasoning+2.9 pts 8Frontier Questions+7.3 pts 6Abstract & Novel Reasoning+5.2 pts 7Logic & Puzzles+2.4 pts 4Planning & Strategy+2.0 pts

Math

5Math+4.1 pts 9Competition Math+4.3 pts 10Research Math+6.3 pts 3Applied & Quantitative+1.7 pts

Science

6Science+2.5 pts 4Graduate Science QA+1.1 pts 4Biology & Medicine+2.5 pts 7Physics & Chemistry+3.2 pts 6Research Assistance+2.4 pts

Writing & Language

5Writing & Language+3.1 pts 6Creative Writing+27 Elo 4Technical Writing+1.7 pts 8Editing & Style+3.3 pts 7Multilingual+1.8 pts 7Instruction Following+3.7 pts 7Conversation & EQ+27 Elo

Multimodal

17Multimodal+15.1 pts 9Image Understanding+6.2 pts 7Documents & Charts+5.5 pts 15Video Understanding+15.0 pts 27Speech & Audio+27.7 pts

Tool Use & Agents

4Tool Use & Agents+2.9 pts 6Tool Use+3.1 pts 7Computer Use+6.6 pts 4Web Research+3.4 pts 4Long-Horizon+9.3 hrs

Long Context

6Long Context+3.3 pts 51M Recall+3.5 pts 6Long Reasoning+2.1 pts

Speed & Cost

53Speed & Cost+56.5 pts 51Output Speed+475 tok/s 52Latency+9.56s 49Cost-Efficiency+56.9 pts

Trust & Factuality

4Trust & Factuality+3.0 pts 4Factuality+2.3 pts 5Honesty & Calibration+3.0 pts

Overall Intelligence over time

#1 frontierFable 5Sonnet 5.5Grok 4.7

Update history

The distilled version first; open any release for the deeper breakdown.

The Claude Fable family

Fable 5June 2026 Fable 5.1August 2026

60-second briefing

Fable 5, explained

AI video summary · placeholder render from the auto-written script

From Anthropic's own guide, distilled

Prompt engineering and Claude best practices

Be explicit about the outcome you want; newer Claude models follow instructions literally rather than guessing at hidden intent.
Explain why a rule exists. Context about the goal lets the model generalise the rule to cases you did not list.
Wrap documents, examples and data in XML-style tags so instructions and material never blur together.
Put long documents at the top of the prompt and your question at the end; recall and grounding both improve.
For agentic tasks, state when to act versus when to ask, and ask for a short plan before large edits.

All official guides, distilled →

About Anthropic

Safety-focused lab building the Claude family of models.

Optimising for

Reliable long-horizon agentic work (especially coding and computer use), honest and calibrated answers, and strong writing, with safety research shipped alongside capability.

San Francisco, US · founded 2021

Closest rivals

Fable 5 is 7th in overall intelligence; leader is Claude Opus 5.5.