Claude Fable 5.2 vs Claude Sonnet 5.5
Wins 4 of 6 areas
Coding · Reasoning · Facts · Math
Wins 1 of 6 areas
Agents
Claude Fable 5.2 is the stronger all-rounder.Claude Sonnet 5.5 is cheaper and better at agents.
Scores updated · 33 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking40Claude Fable 5.24 of 4 tests
- CodingWriting and fixing software30Claude Fable 5.23 of 3 tests
- MathCompetition and research-level math20Claude Fable 5.22 of 2 tests
- FactsGetting facts right instead of making them up10Claude Fable 5.21 of 1 test
- AgentsCarrying out multi-step tasks on its own01Claude Sonnet 5.51 of 1 test
- Long documentsFinding answers in very long texts00Even0 each · 1 tie
Agents, facts and long documents rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Claude Sonnet 5.5 costs 75% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Fable 5.2 pulls ahead
- Solves new puzzle games it has never seenARC-AGI-3+55points ahead
- Short factual questions, answered correctlySimpleQA Verified+47.5points ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+19.5points ahead
Where Claude Sonnet 5.5 pulls ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+4points ahead
Every test, side by side
All 33 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Fable 5.2
- SWE-bench ProClaude Fable 5.2 by 11.79381.3+11.7
- SciCodeClaude Fable 5.2 by 6.767.761+6.7
- SWE-bench MultilingualClaude Fable 5.2 by 6.396.690.3+6.3
ReasoningClaude Fable 5.2
- ARC-AGI-3Claude Fable 5.2 by 5579.324.3+55
- Humanity's Last ExamClaude Fable 5.2 by 19.374.355+19.3
- ARC-AGI-2Claude Fable 5.2 by 8.692.583.9+8.6
- CritPtClaude Fable 5.2 by 435.431.4+4
MathClaude Fable 5.2
- FrontierMath Tier 4Claude Fable 5.2 by 19.510080.5+19.5
- FrontierMath Tiers 1-3 (v2)Claude Fable 5.2 by 6.395.188.8+6.3
Other results21 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- HLE-DiamondClaude Fable 5.2 by 26.980.253.3+26.9
- Terminal-Bench-Science 0.1Claude Fable 5.2 by 2073.353.3+20
- SWE-bench MultimodalClaude Fable 5.2 by 19.874.154.3+19.8
- CritPt v2 HardClaude Fable 5.2 by 17.576.559+17.5
- HLE (with tools)Claude Fable 5.2 by 9.774.264.5+9.7
- OSWorld 2.1 partialClaude Fable 5.2 by 9.289.380.1+9.2
- OfficeQA ProClaude Fable 5.2 by 7.372.965.6+7.3
- OfficeQAClaude Fable 5.2 by 6.583.476.9+6.5
- Program BenchClaude Fable 5.2 by 5.985.679.7+5.9
- HealthBench ProfessionalClaude Fable 5.2 by 4.767.763+4.7
- HealthBench Professional length-adjustedClaude Fable 5.2 by 4.767.763+4.7
- GlobalMMLUClaude Fable 5.2 by 3.99692.1+3.9
- GMMLUClaude Fable 5.2 by 3.99692.1+3.9
- DeepSWE 1.1Claude Fable 5.2 by 3.374.371+3.3
- Terminal-Bench 4.0Claude Sonnet 5.5 by 367.670.6+3
- HealthBenchClaude Sonnet 5.5 by 2.559.662.1+2.5
- HealthBench length-adjustedClaude Sonnet 5.5 by 2.559.662.1+2.5
- MILUClaude Fable 5.2 by 2.594.191.6+2.5
- GDP (Surge AI)Claude Sonnet 5.5 by 1.624.225.8+1.6
- GDPval-AA 2.1Claude Sonnet 5.5 by 13 rating points18311844+13 rating
- ARC-AGI-1Claude Fable 5.2 by 197.596.5+1
Questions people ask
Which is better, Claude Fable 5.2 or Claude Sonnet 5.5?
Claude Fable 5.2 wins four of the six areas where both have results: coding, reasoning, facts and math. Claude Sonnet 5.5 wins agents, and costs 75% less. They are level on long documents.
Which is better for coding?
Claude Fable 5.2. It wins 3 of the 3 coding tests both models report; Claude Sonnet 5.5 wins none.
Which is cheaper?
Claude Fable 5.2 costs $8.00 per million input tokens and $40.00 per million output tokens; Claude Sonnet 5.5 costs $2.00 and $10.00. That makes Claude Sonnet 5.5 about 75% cheaper for the same work.
How do you compare the two?
We use the 33 benchmark tests both models have published scores on. The verdict counts the 12 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 21 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.