SYLIN

Claude Fable

by Anthropic · Model · Language & code models

strong

Top-ranked on most independent leaderboards, but expensive and beaten on some tasks.

#1 agents#1 coding#1 documents#1 image understanding#1 knowledge work#1 robot control#1 web development

Scores 54% on real vs 9% on shuffled protein sequences, a +45.1 point gap, the largest of any LLM tested

AccessAPIPrice$8 in · $40 out per M tokensContext1,000K tokens

Why we say so

Arena puts Fable 5.2 first on agents and web development, Artificial Analysis ranks it first on its Intelligence Index (65 versus Astra's 62), and Vals has it just ahead of Gemini Neon. Xinming Tu found its protein-sequence knowledge far ahead of other LLMs, and Simon Willison built a plugin almost in one shot. Against that, Mercor has Astra well ahead on coding, AI Search saw it fail an image test that Gemini Carbon and DeepSeek Flash passed, and HN commenters call it too pricey for routine coding.

Versions

VersionReleasedRank
Claude Fable 5.2LATEST
strong Current leader on agents, web development and overall index rankings
Nov 9#1 agents#1 coding#1 knowledge work#1 robot control#1 web development#2 3d generation
Claude Fable 5.1Sep 1#1 documents
Claude Fable 5
unproven Anthropic's filter blocked 44% of protein questions, so it was excluded from testing
#1 image understanding

Evidence

  1. Nov 19
    Anthropic document says Fable 5.2 and Mythos 5.2 were capability-capped in training for AI R&D domains like kernel optimization; commenters call it anti-competitive gating
  2. Nov 18
    Beats every open protein language model tested on next-residue prediction (ProGen3-3B +32.5, ProtGPT3-10B +27.7)
  3. Nov 18
    GPT, Gemini and open-weight LLMs score near zero gap; on 56 post-cutoff proteins Fable 5.2 got +22.1, GPT-6.2 Sol -0.4
  4. Nov 15
    #1 on Arena webdev (1824)
  5. Nov 15
    Fable 5.2 (Max) ranks #1 on Agent Arena at +30.4% over 4.7K real sessions, about 1.6x Opus 5.5 (+19.5%); also #1 in Code, Work, Chat
Show all 40
  1. Nov 14
    Claude Fable 5.2 and Mythos 5.2 released; Anthropic's system card acknowledges an unreleased internal model more capable than released ones
  2. Nov 14
    Coded a 90-second history-of-humanity animation from scratch in pure JavaScript, with background music; reviewer calls it better than other frontier models
  3. Nov 12
    Claimed to score much higher than Opus and the best Gemini and GPT models controlling two-armed robots
  4. Nov 11
    On GDPx materials science costs 89% more than Astra with 4.3x output tokens for about the same score; 7.5x Carbon's cost
  5. Nov 11
    One lab's 100 unsaturated environments found Astra and GPT-6.1 Sol produced better solutions; Fable paused to ask a nonexistent human
  6. Nov 11
    #1 on Arena webdev (1838)
  7. Nov 11
    Ranks number one on the Artificial Analysis Intelligence Index, a few points above Gemini 4 Carbon
  8. Nov 11
    Ranks number one on Vals Index for knowledge work, well above the best Gemini models and GPT-6 Astra
  9. Nov 11
    Ranks first on FrontierSWE, but the gap to GPT and Gemini Carbon is within error bars
  10. Nov 11
    Built a detailed third-person Unreal Engine ninja game with a critic-agent loop, taking about seven hours and two five-hour limits; best of frontier models
  11. Nov 11
    Wrote a from-scratch bullet-through-water-balloon physics simulation in 35 minutes with adjustable settings; better than Opus 5.5 or GPT-6 Astra
  12. Nov 11
    Failed Frogbench: could not find the hidden frog even at extra effort; DeepSeek Flash is the only model that finds it
  13. Nov 11
    Identified 2 of 6 brain tumor types correctly, versus 1 of 6 for GPT-6 Astra and Kimi; it answered despite earlier Fable refusing medical questions
  14. Nov 11
    Medical deep-research answer on leukemia was too concise and left out detail; reviewer prefers GPT, DeepSeek and Kimi
  15. Nov 9
    #1 on Arena document (1515)
  16. Nov 9
    #1 on Arena vision (1326)
  17. Oct 19
    #1 on Arena vision (1325)
  18. Sep 27
    Previous best Claude; more expensive and restrictive than Opus 5.5, and ranked just above it on LiveBench
  19. Sep 13
    Work used an unnamed internal research model, described as roughly comparable to Claude Fable 5.1, with dozens of agents in a Claude Code-based harness
  20. Sep 13
    #2 on Arena document (1513)
  21. Sep 6
    Anthropic's best model, pitched for agentic scientific research, coding and knowledge work; beats Fable 5 and GPT 5.6 Sol in its table
  22. Sep 6
    Claimed state of the art on agentic coding, knowledge work, agentic use and Humanity's Last Exam, ahead of GPT Sol and Opus 5
  23. Sep 3
    Anthropic chart: Fable 5.1 beats Fable 5 at every price point on terminal science tasks, clearing 50% versus under 30%
  24. Sep 3
    Autonomously researched earnings and produced a motion-graphics financial video; layout looked a bit better than GLM 5.3 and GPT output
  25. Sep 3
    Directed a matcha commercial via Higgsfield CLI, six clips; video good but audio overlapped. GPT Sol result shown for comparison
  26. Sep 3
    Costs $3.69 per task vs $3.14 for Fable 5, and 4-8x more than GPT, Qwen, Kimi and GLM; evaluating it cost $8,500
  27. Sep 1
    Failed a multilingual pick-up-line test: translations were clean and in native scripts but culturally thin, mostly the same sentence translated
  28. Aug 28
    Scores 21.4% on Terminal-Bench-Science in Claude Code, below Opus 5 (30.0%) and GPT Sol (22.4%)
  29. Aug 22
    Author thinks Fable could finish the watch UI pixel-perfect unsupervised if given a simulator screenshot feedback loop
  30. Aug 22
    Fable 5 (Claude Code, high effort) topped the speedrun: record 2,726, 81.7% of the gap to human record closed, 8.7 agent-days
  31. Aug 20
    One commenter finds Fable 5 less verbose than Opus 5, or better at following style guidelines
  32. Jul 27
    Refused the tumor-identification and atherosclerosis research prompts that Opus 5 answered
  33. Jul 17
    Closed weights, no parameter count published, 1M context, 128K max output, thinking always on; $10 in / $50 out per million tokens
  34. Jul 17
    On the same prompt, Fable 5 produced the most polished UI with 8/8 green checks and responsive tabs; judged second of three
  35. Jun 10
    Took #1 on Arena vision

Cost to run

Alternatives

Claude Opus#2 on Arena agent
GPT AstraRebuilt an Airbnb property in Blender with more faithful details (pillows, robes) than Gemini Carbon and Claude Fable, though exterior was wrong
GPT SolIn Codex Ultra, its Yonkers-to-mech three.js animation was weaker at morphing the city into a mech than Fable 5.2 and Gemini 4 Carbon
KimiRanks above Claude Fable 5.1, GPT-6 and Gemini 4 on Artificial Analysis index at about $2 per task
Gemini CarbonOne commenter found it far ahead of Astra Ultra, Opus 5.5 and Fable 5.1 on their research questions (anecdotal)
GLMSits on the cost-intelligence Pareto frontier at about $0.60 per task; Opus and Fable score higher but cost four to five times more