RTX 4060 Ti × Meta Llama 3.1 8B Instruct
Meta Llama 3.1 8B Instruct on RTX 4060 Ti
49.6best gen tok/s
2,267best prompt tok/s
4,352max context tested
10results
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 | 49.6 | 2,130 | LocalScore ↗ | 512+128 | 645 ms | llamafile 0.9.2 |
| Q4_K - Medium | 49.3 | 2,247 | LocalScore ↗ | 512+128 | 612 ms | llamafile 0.9.1 |
| Q4_K - Medium | 49.1 | 2,267 | LocalScore ↗ | 512+128 | 614 ms | llamafile 0.9.2 |
| Q4_K - Medium | 45.4 | 2,231 | LocalScore ↗ | 512+128 | 618 ms | llamafile 0.9.3 |
| Q4_K - Medium | 45.4 | 2,224 | LocalScore ↗ | 512+128 | 621 ms | llamafile 0.9.3 |
| Q4_K - Medium | 44.9 | 2,182 | LocalScore ↗ | 4,352 | 619 ms | llamafile 0.9.3 |
| Q4_K - Medium | 44.8 | 2,217 | LocalScore ↗ | 512+128 | 621 ms | llamafile 0.9.2 |
| Q4_K - Medium | 42.8 | 1,843 | LocalScore ↗ | 512+128 | 696 ms | llamafile 0.9.3 |
| Q4_K - Medium | 30.7 | 2,112 | LocalScore ↗ | 512+128 | 617 ms | llamafile 0.9.2 |
| Q4_K - Medium | 20.6 | 1,929 | LocalScore ↗ | 512+128 | 665 ms | llamafile 0.9.2 |
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.
What else can RTX 4060 Ti run? →
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.