Gemini 2.5 Pro vs Llama 4 Maverick
Wins 6 of 7 areas
Coding · Agents · Reasoning · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Gemini 2.5 Pro is the stronger all-rounder.Llama 4 Maverick is cheaper.
Scores updated · 29 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.
- ReasoningHard problems that need careful thinking50Gemini 2.5 Pro5 of 5 tests
- CodingWriting and fixing software40Gemini 2.5 Pro4 of 4 tests
- Images and chartsUnderstanding pictures, charts and video30Gemini 2.5 Pro3 of 3 tests
- AgentsCarrying out multi-step tasks on its own10Gemini 2.5 Pro1 of 3 tests · 2 ties
- Long documentsFinding answers in very long texts10Gemini 2.5 Pro1 of 1 test
- Following instructionsDoing exactly what it is asked10Gemini 2.5 Pro1 of 1 test
- FactsGetting facts right instead of making them up11Even1 each
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 90% 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 Gemini 2.5 Pro pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+42.2points ahead
- Common-sense trick questionsSimpleBench+23.9points ahead
- Command-line tasks in a real terminalTerminal-Bench 2.1+20.6points ahead
Where Llama 4 Maverick pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+2points ahead
Every test, side by side
All 29 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGemini 2.5 Pro
- SWE-bench VerifiedGemini 2.5 Pro by 42.263.221+42.2
- Terminal-Bench 2.1Gemini 2.5 Pro by 20.628.57.9+20.6
- Terminal-Bench HardGemini 2.5 Pro by 19.726.56.8+19.7
- SciCodeGemini 2.5 Pro by 14.646.331.7+14.6
AgentsGemini 2.5 Pro
- τ-Bench V3 · BankingGemini 2.5 Pro by 69.73.7+6
- GDPValtie00tie
- Terminal-Bench 4.0tie00tie
ReasoningGemini 2.5 Pro
- SimpleBenchGemini 2.5 Pro by 23.951.627.7+23.9
- GPQA DiamondGemini 2.5 Pro by 16.583.667.1+16.5
- Humanity's Last ExamGemini 2.5 Pro by 13.1184.9+13.1
- ARC-AGI-2Gemini 2.5 Pro by 4.94.90+4.9
- CritPtGemini 2.5 Pro by 2.62.60+2.6
FactsEven
- AA-Omniscience · AccuracyGemini 2.5 Pro by 14.23924.9+14.2
- AA-Omniscience · Non-hallucinationLlama 4 Maverick by 29.111.1+2
Images and chartsGemini 2.5 Pro
Following instructionsGemini 2.5 Pro
- IFBenchGemini 2.5 Pro by 5.748.743+5.7
Other results10 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.
- HMMT Feb. 2025Gemini 2.5 Pro by 72.580.88.3+72.5
- Aider-PolyglotGemini 2.5 Pro by 60.976.515.6+60.9
- AIME 2025Gemini 2.5 Pro by 60.58322.5+60.5
- τ²-Bench Telecom (AA run)Gemini 2.5 Pro by 36.354.117.8+36.3
- ARC-AGI-1Gemini 2.5 Pro by 32.6374.4+32.6
- LiveCodeBenchGemini 2.5 Pro by 30.874.243.4+30.8
- Artificial Analysis Coding IndexGemini 2.5 Pro by 30.446.716.3+30.4
- AA-OmniscienceGemini 2.5 Pro by 25.5-16.4-41.9+25.5
- MMLU-ProGemini 2.5 Pro by 5.58680.5+5.5
- AA IntelligenceGemini 2.5 Pro by 51510+5
Questions people ask
Which is better, Gemini 2.5 Pro or Llama 4 Maverick?
Gemini 2.5 Pro wins six of the seven areas where both have results: coding, agents, reasoning, images and charts, long documents and following instructions. Llama 4 Maverick wins none, but costs 90% less. They are level on facts.
Which is better for coding?
Gemini 2.5 Pro. It wins 4 of the 4 coding tests both models report; Llama 4 Maverick wins none.
Which is cheaper?
Gemini 2.5 Pro costs $1.25 per million input tokens and $10.00 per million output tokens; Llama 4 Maverick costs $0.25 and $0.87. That makes Llama 4 Maverick about 90% cheaper for the same work.
How do you compare the two?
We use the 29 benchmark tests both models have published scores on. The verdict counts the 19 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 10 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.