RTX 2070 × Meta Llama 3.1 8B Instruct
Meta Llama 3.1 8B Instruct on RTX 2070
55.5best gen tok/s
1,316best prompt tok/s
4,352max context tested
5results
Measured · all numbers below come from the linked sources
llama.cpp
| Quant | Gen tok/s | Prompt tok/s | Source | Context tested | TTFT | Build |
|---|---|---|---|---|---|---|
| Q4_K - Medium | 55.5 | 1,316 | LocalScore ↗ | 512+128 | 963 ms | llamafile 0.9.2 |
| Q4_K - Medium | 53.3 | 1,248 | LocalScore ↗ | 4,352 | 1,017 ms | llamafile 0.9.2 |
| Q4_K - Medium | 51.7 | 1,216 | LocalScore ↗ | 512+128 | 1,029 ms | llamafile 0.9.2 |
| Q4_K - Medium | 51.5 | 1,216 | LocalScore ↗ | 4,352 | 1,048 ms | llamafile 0.9.2 |
| Q4_K - Medium | 49.9 | 1,263 | LocalScore ↗ | 4,352 | 1,021 ms | llamafile 0.9.3 |
Ollama
No measured Ollama results yet. Submit yours →
MLX
No measured MLX results yet. Submit yours →
"Context tested" is the largest prompt+generation length actually benchmarked (tokens), not a memory-fit estimate.
Includes results from LocalScore, a Mozilla Builders project by cjpais, mirrored in the cn0303/local-llm-speed-benchmarks dataset. LocalScore runs use llamafile, which is built on llama.cpp.