Each row is a run configuration (model + hardware + framework + quantization); repeated runs of the same setup are merged — raw counts appear in each result’s source line.
NVIDIA RTX 5060 SUPER · 12GB · llama.cpp
Other
Today
145.6 tok/s
Decode
1,092 tok/s
Prefill
0.09 s
TTFT
2.9 GB
VRAM
Sources: Other · 1 runs · 1 independent sources
NVIDIA RTX 5070 · 12GB · llama.cpp
Other
Today
174.7 tok/s
Decode
1,310.3 tok/s
Prefill
0.08 s
TTFT
3.5 GB
VRAM
Sources: Other · 1 runs · 1 independent sources
NVIDIA RTX 3060 12GB · 12GB · llama.cpp
Other
Today
117 tok/s
Decode
877.5 tok/s
Prefill
0.11 s
TTFT
2.9 GB
VRAM
Sources: Other · 1 runs · 1 independent sources
3 results