GPUs & Macs › NVIDIA RTX 5090 Laptop
NVIDIA RTX 5090 Laptop: local LLM speed
Llama 2 7B (Q4_0) generates 158 tokens/s on the NVIDIA RTX 5090 Laptop (CUDA), and processes prompts at 7,642 tokens/s.
Measured results
| Model | Generate tok/s | Prompt tok/s | Quant | Backend | Test | Source |
|---|---|---|---|---|---|---|
| Llama 2 7B | 158 | 7,642 | Q4_0 | CUDA FA on | tg128 | llama.cpp #15013 details |
| Llama 2 7B | 156 | 6,667 | Q4_0 | CUDA FA off | tg128 | llama.cpp #15013 details |
| Llama 2 7B | 154 | 6,011 | Q4_0 | Vulkan FA on | tg128 | llama.cpp #10879 details |
| Llama 2 7B | 149 | 5,546 | Q4_0 | Vulkan FA off | tg128 | llama.cpp #10879 details |
| Llama 3.2 1B Instruct | 160 | 11,638 | Q4_K_M | llamafile | LocalScore avg · n=2 | LocalScore LocalScore details |
| Llama 3.1 8B Instruct | 97.8 | 4,146 | Q4_K_M | llamafile | LocalScore avg · n=2 | LocalScore LocalScore details |
| Qwen2.5 14B Instruct | 40.5 | 1,550 | Q4_K_M | llamafile | LocalScore avg | LocalScore details |
What fits on the NVIDIA RTX 5090 Laptop
Estimates for this chip are scaled from a measured run: Llama 2 7B Q4_0 at 158 tok/s (CUDA, source). Fit assumes the whole model in GPU memory with ~8K context.
| Model | Best quant that fits | File | Expected tok/s |
|---|---|---|---|
| Qwen3.8-27B | UD-Q6_K | 22.0 GB | 27.5 estimate |
| gemma-4-12B-it | Q8_0 | 12.7 GB | 47.7 estimate |
| Qwen3.5-9B | Q8_0 | 9.5 GB | 63.4 estimate |
| Qwen3.5-4B | Q8_0 | 4.5 GB | 135 estimate |
| Qwen3.6-35B-A3B | UD-Q4_K_S | 20.9 GB | 169 estimate |
| gemma-4-26B-A4B-it | UD-Q5_K_M | 21.2 GB | 92.9 estimate |
| MiMo-V2.6-Distill-Qwen-9B | Q8_0 | 9.5 GB | 63.3 estimate |
| gemma-4-E4B-it | Q8_0 | 8.2 GB | 73.8 estimate |
| Muse-Glimmer-30B | UD-Q5_K_M | 19.2 GB | 31.5 estimate |
| Qwen3-4B | Q8_0 | 4.3 GB | 141 estimate |
| gemma-4-31B-it | Q5_K_M | 21.7 GB | 27.9 estimate |
| gemma-4-26B-A4B-it-qat-q4_0-unquantized | UD-Q4_K_XL | 14.2 GB | 138 estimate |
| gemma-4-12B-it-qat-q4_0-unquantized | UD-Q4_K_XL | 6.7 GB | 90.0 estimate |
| gemma-3-1b-it | Q8_0 | 1.1 GB | 504 estimate |
Similar hardware
NVIDIA RTX 3080 Laptop · NVIDIA RTX A5000 Laptop · NVIDIA RTX 5000 Ada Laptop · NVIDIA RTX 4090 Laptop · NVIDIA RTX 5080 Laptop · NVIDIA RTX A4500 Laptop · NVIDIA RTX 5070 Ti Laptop · NVIDIA RTX 3500 Ada Laptop