Related Models
Performance
64
tokens/sec
Faster than 32% of models
0.46
seconds
Faster than 91% of models
0.46
seconds
Faster than 94% of models
Market Median
94 tok/s
32% slower
Median TTFT
1.11s
59% faster
Speed Comparison
Qwen3 Coder 480B A35B Instruct
64 tok/s+0%
Apertus 70B Instruct
64 tok/s+0%
Devstral Medium
64 tok/s+0%
Benchmarks
MMLU-Pro
37.8%
GPQA Diamond
22.9%
HLE
4.0%
LiveCodeBench
9.5%
SciCode
5.2%
TerminalBench Hard
0.8%
MATH-500
69.1%
AIME
9.0%
AIME 2025
10.3%
IFBench
22.0%
Long Context Recall
0.0%
Tau2
0.0%
Market AverageTop Score