RTX A4000 × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on RTX A4000
241best gen tok/s
8,404best prompt tok/s
640max context tested
7results
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 | 241 | 8,404 | LocalScore ↗ | 512+128 | 168 ms | llamafile 0.9.2 |
| Q4_K - Medium | 233 | 8,278 | LocalScore ↗ | 512+128 | 172 ms | llamafile 0.9.2 |
| Q4_K - Medium | 231 | 8,209 | LocalScore ↗ | 512+128 | 175 ms | llamafile 0.9.2 |
| Q4_K - Medium | 213 | 7,200 | LocalScore ↗ | 512+128 | 194 ms | llamafile 0.9.2 |
| Q4_K - Medium | 135 | 7,232 | LocalScore ↗ | 512+128 | 186 ms | llamafile 0.9.3 |
| Q4_K - Medium | 128 | 7,374 | LocalScore ↗ | 512+128 | 185 ms | llamafile 0.9.3 |
| Q4_K - Medium | 118 | 7,455 | LocalScore ↗ | 512+128 | 185 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.
What else can RTX A4000 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.