作者： Jakub Rusinowski · 最后更新： 2026年7月10日
DeepSeek R1 rivals proprietary top-tier models in reasoning and coding tasks using a Mixture-of-Experts architecture. Known for its exceptional logic and math capabilities. Previous generation — superseded by DeepSeek V4, which folds R1's reasoning strengths into a newer MoE architecture. Still widely deployed and supported.
| DeepSeek R1 Distill Llama 8B | Min 6 GB VRAM · Q4_K_M · 128,000 ctx · ollama run deepseek-r1:8b |
| DeepSeek R1 Distill Qwen 32B | Min 20 GB VRAM · Q4_K_M · 128,000 ctx · ollama run deepseek-r1:32b |
| DeepSeek R1 Distill Qwen 14B | Min 9 GB VRAM · Q4_K_M · 128,000 ctx · ollama run deepseek-r1:14b |
| DeepSeek R1 (671B) | Min 406 GB VRAM · Q4_K_M · 128,000 ctx · ollama run deepseek-r1:671b |
The cheapest GPU that runs DeepSeek R1 locally (min 6 GB VRAM) is the Intel Arc B570 (10 GB).
Install Ollama then run: ollama run deepseek-r1:8b
Minimum VRAM: 6 GB. For best results use Q4_K_M quantization.
DeepSeek R1 needs about 6 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: DeepSeek R1 Distill Llama 8B (6 GB, Q4_K_M); DeepSeek R1 Distill Qwen 32B (20 GB, Q4_K_M); DeepSeek R1 Distill Qwen 14B (9 GB, Q4_K_M); DeepSeek R1 (671B) (406 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Yes — DeepSeek R1 runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
Q4_K_M is the best balance of quality and VRAM for DeepSeek R1 in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
Install Ollama, then run: ollama run deepseek-r1:8b. This downloads DeepSeek R1 and starts a local, OpenAI-compatible endpoint — no internet connection is needed after the initial download.