← All models
Compare

Claude Fable 5 vs Claude Sonnet 4.5

Add, remove, or swap models to compare them side by side.

Claude Fable 5Claude Sonnet 4.5
AttributeClaude Fable 5Anthropic · 42 in the newsClaude Sonnet 4.5Anthropic · 4 in the news
Scores
Intelligence (ECI)162147
Coding9931
Math9532
Reasoning & Knowledge9628
Agentic & Tools9827
Specifications
DeveloperAnthropicAnthropic
FamilyClaudeClaude
ReleasedJun 9, 2026Sep 29, 2025
Parameters
AvailabilityAPI accessAPI access
Context window1M1M
Price — $/M input$10.00$3.00
Price — $/M output$50.00$15.00
Inputstext, image, filetext, image, file
Outputstexttext
Benchmarks
AIME 2024/2025100%78%
APEX45%
GPQA Diamond93%82%
Humanity's Last Exam53%14%
SciCode60%
SimpleBench82%54%
SimpleQA Verified68%24%
WebDev Arena16541391
WeirdML92%48%
τ²-bench99%
ARC-AGI64%
ARC-AGI-214%
FrontierMath15%
FrontierMath Tier 44%
GDPval (win/tie rate)50%
GSO (code optimization)15%
MATH Level 598%
METR task horizon2.0 h
SWE-bench Verified71%
Terminal-Bench47%

Use-case scores are 0–100 percentile composites across each area’s benchmarks, ranked against every model from the past year. Highlighted cells lead each row. Open a model for the full picture.

Benchmarks & model data from Epoch AI (CC BY); pricing & specs from OpenRouter. ECI = Epoch Capabilities Index.