Claude Fable 5

Leaderboards

Where Claude Fable 5 sits among the top 20 in each category.

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Test results

Every test Claude Fable 5 has a published result on, out of 100%, with its rank among the models that took it.

• SWE-Bench Verified

  #1 / 123

  Fixing real bugs in real GitHub projects, checked by the projects' own tests.

  95.0%

  Self-reported

• Humanity's Last Exam

  #4 / 115

  Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions

  64.5%

  Self-reported · In the text score· In the vision score

• SWE-Bench Pro

  #3 / 63

  Harder, longer software engineering tasks from real codebases.

  80.0%

  Self-reported · In the coding score

• Terminal-Bench 2.1

  #16 / 53

  Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.

  84.3%

  Self-reported

• DeepSWE 1.1

  #20 / 52

  DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.

  70.0%

  Self-reported · In the coding score

• LiveBench

  #5 / 38

  LiveBench is a challenging, contamination-limited LLM benchmark that addresses test set contamination by releasing new questions monthly based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses. It comprises tasks across math, coding, reasoning, language, instruction following, and data analysis with verifiable, objective ground-truth answers.

  78.3%

  Self-reported · In the text score

• Terminal-Bench 4.0

  #12 / 31

  Terminal-Bench 4.0 evaluates AI agents on real-world work in terminal and command-line containerized environments across 66 tasks, with emphasis on science-adjacent and frontier engineering problems. Relative to earlier Terminal-Bench releases, 4.0 increases timeouts and adaptively raises RAM/CPU on selected tasks to reduce harness and resource confounds.

  44.5%

  Self-reported · In the coding score· In the agents score

• Finance Agent v2

  #4 / 28

  Finance Agent v2 is an agentic financial-analysis benchmark from Vals that evaluates models on real-world finance workflows, measuring their ability to retrieve and reason over financial documents, perform multi-step calculations, and produce accurate analyses.

  56.3%

  Self-reported

• OSWorld-Verified

  #2 / 28

  OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

  85.0%

  Self-reported

• AutomationBench

  #24 / 27

  AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.

  17.4%

  Self-reported · In the agents score

• FrontierCode 1.1

  #2 / 24

  FrontierCode 1.1 evaluates whether coding-agent changes are mergeable, using unit tests, maintainer-defined rubrics, and verifiers. Runs flagged for unfair internet use receive a zero score.

  53.5%

  Self-reported · In the coding score

• HealthBench Professional

  #2 / 20

  HealthBench Professional evaluates model capability and safety for clinician use cases using real clinician-style chats and physician-authored grading rubrics.

  66.0%

  Self-reported

• FrontierSWE

  #1 / 16

  FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. Performance is reported as a dominance score, where higher is better.

  90.0%

  Self-reported

• GDP.pdf

  #11 / 15

  GDP.pdf is a knowledge-work vision benchmark that evaluates models on economically valuable professional tasks presented as visual documents (PDFs), testing document-based reasoning, chart and table interpretation, and problem solving without tools.

  29.8%

  Self-reported

• Legal Agent Benchmark

  #1 / 15

  The Legal Agent Benchmark (LAB) is Harvey's open-source benchmark for evaluating AI agents on complex, long-horizon legal work. Tasks are scored under an all-pass standard against expert-curated rubrics, where a task passes only if every required rubric criterion (facts, conclusions, citations, structure, and analytical moves) passes.

  13.3%

  Self-reported

• ExploitBench

  #2 / 9

  ExploitBench is a cybersecurity benchmark that evaluates a model's ability to discover and exploit software vulnerabilities, reported as the fraction of challenges where the model captures the target (Cap%).

  78.0%

  Self-reported

• BioMysteryBench

  #7 / 8

  BioMysteryBench evaluates a model's ability to reason through challenging molecular biology problems, reporting performance on a hard subset and on the subset of problems solved by human experts.

  46.1%

  Self-reported

• FrontierCode

  #3 / 4

  FrontierCode is Cognition's coding evaluation that tests whether models can pass difficult coding tasks while meeting the standards of high-quality production codebases. The Diamond subset contains the hardest problems.

  46.3%

  Self-reported

• Blueprint-Bench 2

  #2 / 3

  Blueprint-Bench 2 is an agentic spatial reasoning benchmark that evaluates a model's ability to understand, plan, and reason over architectural blueprints and other structured spatial documents. Scores are reported as a normalized score.

  38.6%

  Self-reported