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

bart-base

by facebook

139M params · feature-extraction · 204 likes · 1.8M downloads

bart-base is a 139M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (9417 tok/s), GeForce RTX 5090 (14122 tok/s), M4 Max 128GB (3446 tok/s).

Inference providers

Provider$/1M in$/1M outThroughput
HF Inference

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB9417 tok/scomfortable
GeForce RTX 509032GB14122 tok/scomfortable
M4 Max 128GB128GB3446 tok/scomfortable
M4 Pro 48GB48GB1723 tok/scomfortable
M4 Pro 24GB24GB1723 tok/scomfortable
A100 PCIe 80 GB80GB17254 tok/scomfortable
H100 SXM5 80 GB80GB34703 tok/scomfortable
GeForce RTX 309024GB8326 tok/scomfortable
Radeon RX 7900 XTX24GB6885 tok/scomfortable
GeForce RTX 408016GB6683 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0