GPT-4o vs Llama 4 Maverick
Wins 2 of 7 areas
Coding · Long documents
Wins 3 of 7 areas
Reasoning · Images and charts · Following instructions
Llama 4 Maverick wins more areas, narrowly.GPT-4o is better at coding.
Scores updated · 27 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.
- Images and chartsUnderstanding pictures, charts and video03Llama 4 Maverick3 of 3 tests
- ReasoningHard problems that need careful thinking02Llama 4 Maverick2 of 4 tests · 2 ties
- Following instructionsDoing exactly what it is asked01Llama 4 Maverick1 of 1 test
- CodingWriting and fixing software30GPT-4o3 of 3 tests
- Long documentsFinding answers in very long texts10GPT-4o1 of 1 test
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
- FactsGetting facts right instead of making them up11Even1 each
Agents, long documents and following instructions 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.
Llama 4 Maverick costs 94% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where Llama 4 Maverick pulls ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+14.5points ahead
- Math problems shown in pictures and chartsMathVista+12.3points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+7points ahead
Where GPT-4o pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+51points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+12.2points ahead
- Reasons across sets of long documentsAA-LCR+6points ahead
Every test, side by side
All 27 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-4o
- SWE-bench VerifiedGPT-4o by 12.233.221+12.2
- SciCodeGPT-4o by 1.633.331.7+1.6
- Terminal-Bench HardGPT-4o by 1.58.36.8+1.5
ReasoningLlama 4 Maverick
- GPQA DiamondLlama 4 Maverick by 14.552.667.1+14.5
- Humanity's Last ExamLlama 4 Maverick by 3.11.84.9+3.1
- ARC-AGI-2tie00tie
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationGPT-4o by 5162.111.1+51
- AA-Omniscience · AccuracyLlama 4 Maverick by 1.223.724.9+1.2
Images and chartsLlama 4 Maverick
Following instructionsLlama 4 Maverick
- IFBenchLlama 4 Maverick by 73643+7
Other results12 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-OmniscienceGPT-4o by 31.3-10.5-41.9+31.3
- Aider-PolyglotGPT-4o by 15.130.715.6+15.1
- τ²-Bench Telecom (AA run)GPT-4o by 11.128.917.8+11.1
- AIME 2025Llama 4 Maverick by 10.811.722.5+10.8
- Artificial Analysis Coding IndexGPT-4o by 7.924.216.3+7.9
- MMLU-ProLlama 4 Maverick by 5.874.780.5+5.8
- LiveCodeBenchLlama 4 Maverick by 5.138.343.4+5.1
- ChartQALlama 4 Maverick by 4.385.790+4.3
- AA IntelligenceLlama 4 Maverick by 2.77.310+2.7
- MGSMLlama 4 Maverick by 1.890.592.3+1.8
- DocVQALlama 4 Maverick by 1.692.894.4+1.6
- ARC-AGI-1tie4.54.4tie
Questions people ask
Which is better, GPT-4o or Llama 4 Maverick?
Llama 4 Maverick wins three of the seven areas where both have results: reasoning, images and charts and following instructions. GPT-4o wins coding and long documents. They are level on agents and facts.
Which is better for coding?
GPT-4o. It wins 3 of the 3 coding tests both models report; Llama 4 Maverick wins none.
Which is cheaper?
GPT-4o costs $5.00 per million input tokens and $15.00 per million output tokens; Llama 4 Maverick costs $0.25 and $0.87. That makes Llama 4 Maverick about 94% cheaper for the same work.
How do you compare the two?
We use the 27 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 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.