Claude 3.5 Sonnet wins more areas, narrowly.
Scores updated · 25 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 3.5 Sonnet costs 55% 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 3.5 Sonnet pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+8points ahead
- Common-sense trick questionsSimpleBench+2.4points ahead
Where GPT-4 Turbo pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 25 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningClaude 3.5 Sonnet
- GPQA DiamondClaude 3.5 Sonnet by 85648+8
- SimpleBenchClaude 3.5 Sonnet by 2.427.525.1+2.4
- Humanity's Last Examtie3.23.1tie
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.
- metr_hcastClaude 3.5 Sonnet by 2440.716.7+24
- MMLUClaude 3.5 Sonnet by 17.688.771.1+17.6
- AIR-Bench 2024Claude 3.5 Sonnet by 1485.971.9+14
- GSM8KClaude 3.5 Sonnet by 1496.482.4+14
- MATHGPT-4 Turbo by 12.271.183.3+12.2
- HarmBenchClaude 3.5 Sonnet by 8.398.189.8+8.3
- HumanEvalClaude 3.5 Sonnet by 6.693.787.1+6.6
- metr_re_benchClaude 3.5 Sonnet by 550+5
- Artificial Analysis Coding IndexClaude 3.5 Sonnet by 4.52621.5+4.5
- LiveCodeBenchGPT-4 Turbo by 4.532.837.3+4.5
- MGSMClaude 3.5 Sonnet by 3.191.688.5+3.1
- XSTestGPT-4 Turbo by 2.195.697.7+2.1
- naturalquestions_closedbookClaude 3.5 Sonnet by 250.248.2+2
- NarrativeQAGPT-4 Turbo by 1.574.676.1+1.5
- DROPClaude 3.5 Sonnet by 1.187.186+1.1
- simple_safety_testsClaude 3.5 Sonnet by 110099+1
- bbqtie94.994.1tie
- metr_swaatie98.799.5tie
- AA Intelligencetie7.27tie
- OpenBookQAtie97.297tie
- anthropic_red_teamtie99.899.7tie
Questions people ask
Which is better, Claude 3.5 Sonnet or GPT-4 Turbo?
Claude 3.5 Sonnet wins one of the two areas where both have results: reasoning. GPT-4 Turbo wins none. They are level on coding.
Which is better for coding?
Neither. All 1 coding tests both models report are ties.
Which is cheaper?
Claude 3.5 Sonnet costs $3.00 per million input tokens and $15.00 per million output tokens; GPT-4 Turbo costs $10.00 and $30.00. That makes Claude 3.5 Sonnet about 55% cheaper for the same work.
How do you compare the two?
We use the 25 benchmark tests both models have published scores on. The verdict counts the 4 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.