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

MiniCPM5-1B

by openbmb

1.1B params · text-generation · 813 likes · 280.8k downloads

MiniCPM5-1B is a 1.1B parameter model. At Q4 quantization it requires 1GB of VRAM. It runs comfortably on GeForce RTX 4090 (1215 tok/s), GeForce RTX 5090 (1822 tok/s), M4 Max 128GB (444 tok/s).

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB1215 tok/scomfortable
GeForce RTX 509032GB1822 tok/scomfortable
M4 Max 128GB128GB444 tok/scomfortable
M4 Pro 48GB48GB222 tok/scomfortable
M4 Pro 24GB24GB222 tok/scomfortable
A100 PCIe 80 GB80GB2226 tok/scomfortable
H100 SXM5 80 GB80GB4477 tok/scomfortable
GeForce RTX 309024GB1074 tok/scomfortable
Radeon RX 7900 XTX24GB888 tok/scomfortable
GeForce RTX 408016GB862 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0