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

gemma-3-270m-it

by google

268M params · text-generation · 595 likes · 135.5k downloads

gemma-3-270m-it is a 268M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (4897 tok/s), GeForce RTX 5090 (7344 tok/s), M4 Max 128GB (1792 tok/s).

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB4897 tok/scomfortable
GeForce RTX 509032GB7344 tok/scomfortable
M4 Max 128GB128GB1792 tok/scomfortable
M4 Pro 48GB48GB896 tok/scomfortable
M4 Pro 24GB24GB896 tok/scomfortable
A100 PCIe 80 GB80GB8972 tok/scomfortable
H100 SXM5 80 GB80GB18047 tok/scomfortable
GeForce RTX 309024GB4330 tok/scomfortable
Radeon RX 7900 XTX24GB3580 tok/scomfortable
GeForce RTX 408016GB3475 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0