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

Nanbeige4.1-3B

by Nanbeige

3.9B params · text-generation · 992 likes · 567.7k downloads

Nanbeige4.1-3B is a 3.9B parameter model. At Q4 quantization it requires 2GB of VRAM. It runs comfortably on GeForce RTX 4090 (333 tok/s), GeForce RTX 5090 (500 tok/s), M4 Max 128GB (122 tok/s).

Inference providers

Provider$/1M in$/1M outThroughput
Featherless

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB333 tok/scomfortable
GeForce RTX 509032GB500 tok/scomfortable
M4 Max 128GB128GB122 tok/scomfortable
M4 Pro 48GB48GB61 tok/scomfortable
M4 Pro 24GB24GB61 tok/scomfortable
A100 PCIe 80 GB80GB611 tok/scomfortable
H100 SXM5 80 GB80GB1230 tok/scomfortable
GeForce RTX 309024GB295 tok/scomfortable
Radeon RX 7900 XTX24GB244 tok/scomfortable
GeForce RTX 408016GB236 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0