GPT-4 Turbo wins more areas, narrowly.Claude 3 Haiku is cheaper.
Scores updated · 22 tests both models report · How we compare
Where each one wins
Tests won in each of the two 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.
Coding 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 3 Haiku costs 96% 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 GPT-4 Turbo pulls ahead
- Code for real scientific research problemsSciCode+13.3points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+10.6points ahead
Where Claude 3 Haiku pulls ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+1points ahead
Every test, side by side
All 22 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningEven
- GPQA DiamondGPT-4 Turbo by 10.637.448+10.6
- Humanity's Last ExamClaude 3 Haiku by 14.13.1+1
Other results19 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.
- NarrativeQAGPT-4 Turbo by 51.724.476.1+51.7
- naturalquestions_closedbookGPT-4 Turbo by 33.814.448.2+33.8
- livecodebench_mediumGPT-4 Turbo by 21.7930.7+21.7
- livecodebench_easyGPT-4 Turbo by 20.361.181.4+20.3
- Artificial Analysis Coding IndexGPT-4 Turbo by 14.86.721.5+14.8
- LiveCodeBenchGPT-4 Turbo by 14.822.537.3+14.8
- MGSMGPT-4 Turbo by 13.475.188.5+13.4
- OpenBookQAGPT-4 Turbo by 13.283.897+13.2
- XSTestGPT-4 Turbo by 12.485.397.7+12.4
- HumanEvalGPT-4 Turbo by 11.275.987.1+11.2
- AIR-Bench 2024Claude 3 Haiku by 10.882.771.9+10.8
- DROPGPT-4 Turbo by 7.678.486+7.6
- GSM8KClaude 3 Haiku by 6.588.982.4+6.5
- MMLUGPT-4 Turbo by 4.966.271.1+4.9
- livecodebench_hardGPT-4 Turbo by 2.11.94+2.1
- HarmBenchClaude 3 Haiku by 1.591.389.8+1.5
- AA IntelligenceGPT-4 Turbo by 1.45.67+1.4
- simple_safety_testsClaude 3 Haiku by 110099+1
- anthropic_red_teamtie10099.7tie
Questions people ask
Which is better, Claude 3 Haiku or GPT-4 Turbo?
GPT-4 Turbo wins one of the two areas where both have results: coding. Claude 3 Haiku wins none, but costs 96% less. They are level on reasoning.
Which is better for coding?
GPT-4 Turbo. It wins the one coding test both models report.
Which is cheaper?
Claude 3 Haiku costs $0.25 per million input tokens and $1.25 per million output tokens; GPT-4 Turbo costs $10.00 and $30.00. That makes Claude 3 Haiku about 96% cheaper for the same work.
How do you compare the two?
We use the 22 benchmark tests both models have published scores on. The verdict counts the 3 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 19 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.