DeepSeek-R1
#124 of 274 scored.
DeepSeekJan 20, 2025Open weights
Capabilities Index
Range 136.3–140.3
139.0
Context
Takes text
64K
Input
Per 1M tokens
$0.70
Output
Per 1M tokens
$2.50
Benchmarks
How it scores.
- GPQA Diamond62.3%
98th of 186
- ARC-AGI-21.3%
62nd of 79
- Mock AIME53.3%
112th of 176
Everything else Epoch has for it
- MATH level 593.0%
- Lech Mazur Writing83.0%
- Fiction.LiveBench69.4%
- Aider polyglot56.9%
- METR Time Horizons51.9%
- WeirdML36.5%
- Balrog34.9%
- SimpleBench17.1%
- ARC-AGI15.8%
On the index
Around it.
Sources