vram.run Models Hardware Providers Cloud State of Inference
API provider data is live · Hardware & cloud pricing curated 2026-02-23

DeepSeek-Coder-V2-Lite-Instruct

by deepseek-ai

15.7B params · text-generation · 557 likes · 205.0k downloads

DeepSeek-Coder-V2-Lite-Instruct is a 15.7B parameter model. At Q4 quantization it requires 8GB of VRAM. It runs comfortably on GeForce RTX 4090 (83 tok/s), GeForce RTX 5090 (125 tok/s), M4 Max 128GB (30 tok/s).

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB83 tok/scomfortable
GeForce RTX 509032GB125 tok/scomfortable
M4 Max 128GB128GB30 tok/scomfortable
M4 Pro 48GB48GB15 tok/stight
M4 Pro 24GB24GB15 tok/stight
A100 PCIe 80 GB80GB153 tok/scomfortable
H100 SXM5 80 GB80GB308 tok/scomfortable
GeForce RTX 309024GB73 tok/scomfortable
Radeon RX 7900 XTX24GB61 tok/scomfortable
GeForce RTX 408016GB59 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0