What Do You Need to Run DeepSeek Locally? VRAM by Version
DeepSeek currently has 6 versions you can deploy locally. The smallest, DeepSeek R1 Qwen3 8B, needs 7GB of VRAM at Q4; the largest, DeepSeek V4 Pro, needs 17GB. The cheapest device that gets one running is the Tesla V100 16GB, at about $305.
VRAM needed for each DeepSeek version
eBay
| Version | Released | Context | |||
|---|---|---|---|---|---|
| DeepSeek R1 Qwen3 8BDense | 8.2B | May 29, 2025 | 128K | 7 | Tesla V100 16GB$305 · 16 GB |
| DeepSeek R1 Distill Qwen 14BDense | 14.8B | Jan 20, 2025 | 128K | 11 | Tesla V100 16GB$305 · 16 GB |
| DeepSeek R1 Distill Qwen 32BDense | 32.8B | Jan 20, 2025 | 128K | 22 | GeForce RTX 2080 Ti 22GB$551 · 22 GB |
| DeepSeek V4 FlashMoE | 284B (13B active) | Jul 31, 2026 | 1M | 7plus 155 GB of RAM | Tesla V100 16GB$305 · 16 GB · Experts offloaded |
| DeepSeek V3.2MoE | 671B (37B active) | Dec 1, 2025 | 160K | 12plus 365 GB of RAM | Tesla V100 16GB$305 · 16 GB · Experts offloaded |
| DeepSeek V4 ProMoE | 1,600B (49B active) | Aug 13, 2026 | 1M | 17plus 879 GB of RAM | GeForce RTX 2080 Ti 22GB$551 · 22 GB · Experts offloaded |
Minimum VRAM = Q4_K_M weights + KV cache for an 8K context + 1GB runtime overhead, rounded up. The cheapest device is the first one that fits when sorting by median price on the current price basis.
About DeepSeek
DeepSeek is published by DeepSeek under MIT. This page covers 6 versions ranging from 8.2B to 1,600B parameters. New official releases are scanned weekly; architecture parameters and sourced or estimated weight sizes are reviewed before publication.
VendorDeepSeek
LicenseMIT
Official sitehttps://www.deepseek.com/
Hugging Face orgdeepseek-ai