Runs on Data centre · Hugging Face ↗
Coding index 69.1 (Artificial Analysis)
The lighter DeepSeek V4 — a 158B MoE, far more self-hostable than the 862B V4 Pro while still a capable open coder. MIT-licensed, 1M context.
Real GGUF file sizes = weight VRAM. Add the KV cache for your context (≈3 GB at 32K, fp16).
| Quantization | Size |
|---|---|
| Q8_0 | 10.9 GB |
| BF16 | 11.3 GB |
| Q2_K | 96.8 GB |
| Q3_K_M | 128.1 GB |
| IQ4_XS | 136.7 GB |
| Q4 | 155.1 GB |
| Q8 | 161.9 GB |
~138 GB at Q4 — a 2×80 GB node, or heavy CPU offload with llama.cpp on less. The most accessible of the strong DeepSeek line for local use.