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

MiniMax-M2.1

by MiniMaxAI

229B params · text-generation · 1.3k likes · 64.5k downloads

MiniMax-M2.1 is a 229B parameter model. At Q4 quantization it requires 114GB of VRAM. It runs comfortably on H200 SXM 141 GB (30 tok/s).

Inference providers

Provider$/1M in$/1M outThroughput
Novita27 tok/s
Fireworks84 tok/s

GPU compatibility

GPUVRAMQ4 DecodeVerdict
H200 SXM 141 GB141GB30 tok/scomfortable
M2 Ultra 192GB192GB2 tok/stight
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0