Prefill Throughput on Synthetic
20,864.85Prefill Throughput (tokens/s)YOCO (Dense)
Evaluation Results
| Method | Links | |
|---|---|---|
| YOCO (Dense)Context Length=128K, Hardware=NVIDIA B200 GPUs2026.06 | 20,864.85 | |
| YOCO (CLSA)Context Length=128K, Hardware=NVIDIA B200 GPUs2026.06 | 20,741.51 | |
| YOCO (CLSA)Context Length=64K, Hardware=NVIDIA B200 GPUs2026.06 | 20,349.11 | |
| YOCO (Dense)Context Length=64K, Hardware=NVIDIA B200 GPUs2026.06 | 20,343.94 | |
| YOCO (Dense)Context Length=32K, Hardware=NVIDIA B200 GPUs2026.06 | 18,450.18 | |
| YOCO (CLSA)Context Length=32K, Hardware=NVIDIA B200 GPUs2026.06 | 18,163.24 | |
| YOCO (Dense)Context Length=16K, Hardware=NVIDIA B200 GPUs2026.06 | 13,033.56 | |
| YOCO (CLSA)Context Length=16K, Hardware=NVIDIA B200 GPUs2026.06 | 11,572.4 | |
| YOCO (Dense)Context Length=8K, Hardware=NVIDIA B200 GPUs2026.06 | 10,884.35 | |
| YOCO (CLSA)Context Length=8K, Hardware=NVIDIA B200 GPUs2026.06 | 9,623.48 | |
| TransformerContext Length=8K, Hardware=NVIDIA B200 GPUs2026.06 | 4,721.71 | |
| TransformerContext Length=16K, Hardware=NVIDIA B200 GPUs2026.06 | 4,160.92 | |
| TransformerContext Length=32K, Hardware=NVIDIA B200 GPUs2026.06 | 2,889.39 | |
| TransformerContext Length=64K, Hardware=NVIDIA B200 GPUs2026.06 | 1,741.56 | |
| TransformerContext Length=128K, Hardware=NVIDIA B200 GPUs2026.06 | 1,019.06 |