Claude Sonnet 4.5 vs o3
Wins 2 of 8 areas
Coding · Reasoning
Wins 4 of 8 areas
Facts · Images and charts · Math · Following instructions
o3 wins more areas, narrowly.Claude Sonnet 4.5 is better at reasoning.
Scores updated · 42 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- Images and chartsUnderstanding pictures, charts and video04o34 of 4 tests
- Following instructionsDoing exactly what it is asked02o32 of 2 tests
- FactsGetting facts right instead of making them up12o32 of 3 tests
- MathCompetition and research-level math01o31 of 1 test
- ReasoningHard problems that need careful thinking21Claude Sonnet 4.52 of 5 tests · 2 ties
- CodingWriting and fixing software21Claude Sonnet 4.52 of 3 tests
- AgentsCarrying out multi-step tasks on its own11Even1 each
- Long documentsFinding answers in very long texts11Even1 each
Math rests on a single test.
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.
o3 costs 44% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where o3 pulls ahead
- Finds hard-to-locate facts by browsing the webBrowseComp+25.6points ahead
- Short factual questions, answered correctlySimpleQA Verified+18.7points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+14.1points ahead
Where Claude Sonnet 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+39points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+8.1points ahead
- Real work tasks from 44 professionsGDPVal+7.7points ahead
Every test, side by side
All 42 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4.5
- SWE-bench VerifiedClaude Sonnet 4.5 by 8.177.269.1+8.1
- SciCodeClaude Sonnet 4.5 by 4.745.741+4.7
- Terminal-Bench Hardo3 by 1.535.637.1+1.5
AgentsEven
- BrowseCompo3 by 25.624.149.7+25.6
- GDPValClaude Sonnet 4.5 by 7.720.512.8+7.7
ReasoningClaude Sonnet 4.5
- ARC-AGI-2Claude Sonnet 4.5 by 7.113.66.5+7.1
- Humanity's Last Examo3 by 2.317.820.1+2.3
- SimpleBenchClaude Sonnet 4.5 by 1.254.353.1+1.2
- GPQA Diamondtie83.482.7tie
- CritPttie1.11.1tie
Factso3
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 3950.911.9+39
- SimpleQA Verifiedo3 by 18.730.749.4+18.7
- AA-Omniscience · Accuracyo3 by 5.732.938.6+5.7
Images and chartso3
- CharXiv (RQ)o3 by 11.467.278.6+11.4
- MathVistao3 by 779.886.8+7
- MMMUo3 by 3.379.682.9+3.3
- MMMU-Proo3 by 1.468.770.1+1.4
Long documentsEven
- LongBench v2Claude Sonnet 4.5 by 361.858.8+3
- AA-LCRo3 by 2.472.374.7+2.4
Following instructionso3
- IFBencho3 by 14.157.371.4+14.1
- Multi-Challengeo3 by 5.155.360.4+5.1
Other results20 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.
- AA Agentic Indexo3 by 18.617.536.1+18.6
- TAU-bench (airline)Claude Sonnet 4.5 by 187052+18
- AA-OmniscienceClaude Sonnet 4.5 by 15.4-0.1-15.6+15.4
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 13.752.138.4+13.7
- LiveCodeBencho3 by 13.77184.7+13.7
- TAU-bench (retail)Claude Sonnet 4.5 by 12.386.273.9+12.3
- HMMT Feb. 2025o3 by 1067.577.5+10
- HarmBencho3 by 6.49298.4+6.4
- τ²-Bench (Retail)Claude Sonnet 4.5 by 686.280.2+6
- AIR-Bench 2024Claude Sonnet 4.5 by 5.389.884.5+5.3
- ARC-AGI-1Claude Sonnet 4.5 by 2.963.760.8+2.9
- τ²-Bench Telecom (AA run)o3 by 2.678.180.7+2.6
- MMLU-ProClaude Sonnet 4.5 by 2.587.585+2.5
- frontiermath_tier_4_v1Claude Sonnet 4.5 by 2.14.22.1+2.1
- bbqClaude Sonnet 4.5 by 198.997.9+1
- simple_safety_testsClaude Sonnet 4.5 by 110099+1
- XSTesttie96.497.3tie
- AIME 2025tie8786.4tie
- AA Intelligencetie20.720.2tie
- anthropic_red_teamtie98.298.3tie
Questions people ask
Which is better, Claude Sonnet 4.5 or o3?
o3 wins four of the eight areas we test: facts, images and charts, math and following instructions. Claude Sonnet 4.5 wins coding and reasoning. They are level on agents and long documents.
Which is better for coding?
Claude Sonnet 4.5. It wins 2 of the 3 coding tests both models report; o3 wins 1.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; o3 costs $2.00 and $8.00. That makes o3 about 44% cheaper for the same work.
How do you compare the two?
We use the 42 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 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.