Claude Sonnet 4 vs GPT-4o
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Claude Sonnet 4 is the stronger all-rounder.
Scores updated · 53 tests both models report · How we compare
Where each one wins
Tests won in each of the seven 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.
- CodingWriting and fixing software40Claude Sonnet 44 of 4 tests
- ReasoningHard problems that need careful thinking30Claude Sonnet 43 of 4 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own20Claude Sonnet 42 of 2 tests
- Images and chartsUnderstanding pictures, charts and video20Claude Sonnet 42 of 2 tests
- Following instructionsDoing exactly what it is asked20Claude Sonnet 42 of 2 tests
- Long documentsFinding answers in very long texts10Claude Sonnet 41 of 1 test
- FactsGetting facts right instead of making them up11Even1 each · 1 tie
Long documents 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.
Claude Sonnet 4 costs 10% 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 pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+39.5points ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+37.2points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+25.1points ahead
Where GPT-4o pulls ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+1points ahead
Every test, side by side
All 53 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Sonnet 4
- SWE-bench VerifiedClaude Sonnet 4 by 39.572.733.2+39.5
- Terminal-Bench HardClaude Sonnet 4 by 22.831.18.3+22.8
- LiveCodeBench v6Claude Sonnet 4 by 17.648.530.9+17.6
- SciCodeClaude Sonnet 4 by 6.74033.3+6.7
AgentsClaude Sonnet 4
- OSWorld-VerifiedClaude Sonnet 4 by 37.242.25+37.2
- GDPValClaude Sonnet 4 by 9.79.70+9.7
ReasoningClaude Sonnet 4
- GPQA DiamondClaude Sonnet 4 by 25.177.752.6+25.1
- Humanity's Last ExamClaude Sonnet 4 by 8.910.71.8+8.9
- ARC-AGI-2Claude Sonnet 4 by 5.95.90+5.9
- CritPttie0.30tie
FactsEven
- AA-Omniscience · Non-hallucinationClaude Sonnet 4 by 8.870.962.1+8.8
- AA-Omniscience · AccuracyGPT-4o by 122.723.7+1
- Vectara HHEM hallucination ratelower is bettertie10.39.6tie
Images and chartsClaude Sonnet 4
Following instructionsClaude Sonnet 4
- IFBenchClaude Sonnet 4 by 18.754.736+18.7
- Multi-ChallengeClaude Sonnet 4 by 16.857.140.3+16.8
Other results35 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.
- AIME 2025Claude Sonnet 4 by 58.870.511.7+58.8
- CNMO 2024Claude Sonnet 4 by 49.660.410.8+49.6
- livecodebench_mediumClaude Sonnet 4 by 43.97632.1+43.9
- Aider-PolyglotClaude Sonnet 4 by 4070.730.7+40
- τ²-Bench Telecom (AA run)Claude Sonnet 4 by 35.764.628.9+35.7
- ARC-AGI-1Claude Sonnet 4 by 35.5404.5+35.5
- AIME 2024Claude Sonnet 4 by 34.143.49.3+34.1
- LiveCodeBenchClaude Sonnet 4 by 30.268.538.3+30.2
- livecodebench_hardClaude Sonnet 4 by 28.132.64.5+28.1
- AIR-Bench 2024Claude Sonnet 4 by 25.988.362.4+25.9
- SimpleQAGPT-4o by 22.315.938.2+22.3
- TAU-bench (retail)Claude Sonnet 4 by 20.280.560.3+20.2
- MATH-500 (EM)Claude Sonnet 4 by 19.49474.6+19.4
- MMLUClaude Sonnet 4 by 17.791.573.8+17.7
- TAU-bench (airline)Claude Sonnet 4 by 17.26042.8+17.2
- livecodebench_easyClaude Sonnet 4 by 16.398.882.5+16.3
- HarmBenchClaude Sonnet 4 by 15.19882.9+15.1
- Artificial Analysis Coding IndexClaude Sonnet 4 by 13.437.624.2+13.4
- SuperGPQAClaude Sonnet 4 by 13.355.742.4+13.3
- AA IntelligenceClaude Sonnet 4 by 11.618.97.3+11.6
- τ²-Bench (Retail)Claude Sonnet 4 by 11.67563.4+11.6
- AA-OmniscienceClaude Sonnet 4 by 10.70.2-10.5+10.7
- Tau2 airlineClaude Sonnet 4 by 1055.545.5+10
- AA Agentic IndexClaude Sonnet 4 by 9.217.68.4+9.2
- MMLU-ProClaude Sonnet 4 by 983.774.7+9
- IFEvalClaude Sonnet 4 by 6.687.681+6.6
- vectara_avg_summary_lengthClaude Sonnet 4 by 59.2 rating points145.886.6+59.2 rating
- MMLU-ReduxClaude Sonnet 4 by 5.693.688+5.6
- MMMLUClaude Sonnet 4 by 5.186.581.4+5.1
- vectara_answer_rateClaude Sonnet 4 by 4.898.693.8+4.8
- bbqClaude Sonnet 4 by 1.596.695.1+1.5
- simple_safety_testsClaude Sonnet 4 by 1.510098.5+1.5
- vectara_factual_consistencytie89.790.4tie
- XSTesttie96.697.3tie
- anthropic_red_teamtie99.299.1tie
Questions people ask
Which is better, Claude Sonnet 4 or GPT-4o?
Claude Sonnet 4 wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. GPT-4o wins none. They are level on facts.
Which is better for coding?
Claude Sonnet 4. It wins 4 of the 4 coding tests both models report; GPT-4o wins none.
Which is cheaper?
Claude Sonnet 4 costs $3.00 per million input tokens and $15.00 per million output tokens; GPT-4o costs $5.00 and $15.00. That makes Claude Sonnet 4 about 10% cheaper for the same work.
How do you compare the two?
We use the 53 benchmark tests both models have published scores on. The verdict counts the 18 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 35 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.