GTX 1660 SUPER × Llama 3.2 1B Instruct
Llama 3.2 1B Instruct on GTX 1660 SUPER
36.5best gen tok/s
825best 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 | 36.5 | 825 | LocalScore ↗ | 512+128 | 1,755 ms | llamafile 0.9.2 |
| Q4_K - Medium | 35.5 | 793 | LocalScore ↗ | 4,352 | 1,838 ms | llamafile 0.9.2 |
| Q4_K - Medium | 35.3 | 793 | LocalScore ↗ | 4,352 | 1,803 ms | llamafile 0.9.2 |
| Q4_K - Medium | 35.3 | 801 | LocalScore ↗ | 4,352 | 1,812 ms | llamafile 0.9.2 |
| Q4_K - Medium | 35.2 | 789 | LocalScore ↗ | 4,352 | 1,824 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 GTX 1660 SUPER 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.