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

biomedical-ner-all

by d4data

66M params · token-classification · 186 likes · 69.0k downloads

biomedical-ner-all is a 66M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (19765 tok/s), GeForce RTX 5090 (29641 tok/s), M4 Max 128GB (7233 tok/s).

Inference providers

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

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB19765 tok/scomfortable
GeForce RTX 509032GB29641 tok/scomfortable
M4 Max 128GB128GB7233 tok/scomfortable
M4 Pro 48GB48GB3616 tok/scomfortable
M4 Pro 24GB24GB3616 tok/scomfortable
A100 PCIe 80 GB80GB36213 tok/scomfortable
H100 SXM5 80 GB80GB72837 tok/scomfortable
GeForce RTX 309024GB17476 tok/scomfortable
Radeon RX 7900 XTX24GB14451 tok/scomfortable
GeForce RTX 408016GB14027 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0