Llama 4 Maverick
#152 of 274 scored.
MetaApr 6, 2025Open weights
Capabilities Index
Range 127.7–134.3
132.2
Context
Takes text, image
1.05M
Input
Per 1M tokens
$0.19
Output
Per 1M tokens
$0.65
Benchmarks
How it scores.
- GPQA Diamond56.0%
105th of 186
- ARC-AGI-20.0%
72nd of 79
- Humanity's Last Exam0.9%
33rd of 40
- Mock AIME20.5%
128th of 176
Everything else Epoch has for it
- MATH level 573.0%
- Lech Mazur Writing62.0%
- GeoBench52.0%
- Fiction.LiveBench46.2%
- DTBench36.4%
- WeirdML24.5%
- LMCA18.7%
- Aider polyglot15.6%
- SimpleBench13.2%
- ARC-AGI4.4%
- FrontierMath-2025-02-28-Private1.2%
On the index
Around it.
Sources