Claude Sonnet 4.5 vs GPT-4.1
Wins 8 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Long documents · Following instructions
Wins 0 of 8 areas
—
Claude Sonnet 4.5 is the stronger all-rounder.GPT-4.1 is cheaper.
Scores updated · 43 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.
- CodingWriting and fixing software40Claude Sonnet 4.54 of 4 tests
- ReasoningHard problems that need careful thinking40Claude Sonnet 4.54 of 4 tests
- Images and chartsUnderstanding pictures, charts and video40Claude Sonnet 4.54 of 4 tests
- Following instructionsDoing exactly what it is asked20Claude Sonnet 4.52 of 2 tests
- FactsGetting facts right instead of making them up21Claude Sonnet 4.52 of 4 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own10Claude Sonnet 4.51 of 1 test
- MathCompetition and research-level math10Claude Sonnet 4.51 of 1 test
- Long documentsFinding answers in very long texts10Claude Sonnet 4.51 of 1 test
Agents, math 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.
GPT-4.1 costs 44% 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 Sonnet 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+44.2points ahead
- Common-sense trick questionsSimpleBench+27.3points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+22.6points ahead
Where GPT-4.1 pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 43 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 22.677.254.6+22.6
- Terminal-Bench HardClaude Sonnet 4.5 by 2235.613.6+22
- LiveCodeBench v6Claude Sonnet 4.5 by 19.36444.7+19.3
- SciCodeClaude Sonnet 4.5 by 7.645.738.1+7.6
ReasoningClaude Sonnet 4.5
- SimpleBenchClaude Sonnet 4.5 by 27.354.327+27.3
- GPQA DiamondClaude Sonnet 4.5 by 16.883.466.6+16.8
- Humanity's Last ExamClaude Sonnet 4.5 by 13.617.84.2+13.6
- CritPtClaude Sonnet 4.5 by 1.11.10+1.1
FactsClaude Sonnet 4.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 4.5 by 44.250.96.7+44.2
- Vectara HHEM hallucination ratelower is betterGPT-4.1 by 6.4125.6+6.4
- AA-Omniscience · AccuracyClaude Sonnet 4.5 by 5.132.927.8+5.1
- SimpleQA Verifiedtie30.731.1tie
Images and chartsClaude Sonnet 4.5
- CharXiv (RQ)Claude Sonnet 4.5 by 10.567.256.7+10.5
- MathVistaClaude Sonnet 4.5 by 7.679.872.2+7.6
- MMMU-ProClaude Sonnet 4.5 by 7.568.761.2+7.5
- MMMUClaude Sonnet 4.5 by 4.879.674.8+4.8
MathClaude Sonnet 4.5
- FrontierMath Tiers 1-3 (v2)Claude Sonnet 4.5 by 17.923.96+17.9
Following instructionsClaude Sonnet 4.5
- Multi-ChallengeClaude Sonnet 4.5 by 1755.338.3+17
- IFBenchClaude Sonnet 4.5 by 14.357.343+14.3
Other results22 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.
- ARC-AGI-1Claude Sonnet 4.5 by 58.263.75.5+58.2
- HMMT 2025Claude Sonnet 4.5 by 45.774.628.9+45.7
- AIME 2025Claude Sonnet 4.5 by 40.68746.4+40.6
- AA-OmniscienceClaude Sonnet 4.5 by 39.5-0.1-39.6+39.5
- τ²-Bench Telecom (AA run)Claude Sonnet 4.5 by 3178.147.1+31
- Artificial Analysis Coding IndexClaude Sonnet 4.5 by 30.352.121.8+30.3
- AIR-Bench 2024Claude Sonnet 4.5 by 2589.864.8+25
- TAU-bench (airline)Claude Sonnet 4.5 by 20.67049.4+20.6
- Terminal-BenchClaude Sonnet 4.5 by 19.75030.3+19.7
- TAU-bench (retail)Claude Sonnet 4.5 by 18.286.268+18.2
- AA Agentic IndexGPT-4.1 by 9.817.527.3+9.8
- AA IntelligenceClaude Sonnet 4.5 by 820.712.7+8
- vectara_factual_consistencyGPT-4.1 by 6.48894.4+6.4
- bbqClaude Sonnet 4.5 by 6.398.992.6+6.3
- vectara_answer_rateGPT-4.1 by 4.395.699.9+4.3
- frontiermath_tier_4_v1Claude Sonnet 4.5 by 4.24.20+4.2
- vectara_avg_summary_lengthClaude Sonnet 4.5 by 36.1 rating points127.891.7+36.1 rating
- MMMLUClaude Sonnet 4.5 by 1.889.187.3+1.8
- XSTestGPT-4.1 by 1.596.497.9+1.5
- anthropic_red_teamGPT-4.1 by 1.198.299.3+1.1
- HarmBenchtie9291.7tie
- simple_safety_teststie100100tie
Questions people ask
Which is better, Claude Sonnet 4.5 or GPT-4.1?
Claude Sonnet 4.5 wins all eight areas we test: coding, agents, reasoning, facts, images and charts, math, long documents and following instructions. GPT-4.1 wins none, but costs 44% less.
Which is better for coding?
Claude Sonnet 4.5. It wins 4 of the 4 coding tests both models report; GPT-4.1 wins none.
Which is cheaper?
Claude Sonnet 4.5 costs $3.00 per million input tokens and $15.00 per million output tokens; GPT-4.1 costs $2.00 and $8.00. That makes GPT-4.1 about 44% cheaper for the same work.
How do you compare the two?
We use the 43 benchmark tests both models have published scores on. The verdict counts the 21 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.