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

wikineural-multilingual-ner

by Babelscape

177M params · token-classification · 160 likes · 434.9k downloads

wikineural-multilingual-ner is a 177M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (7406 tok/s), GeForce RTX 5090 (11107 tok/s), M4 Max 128GB (2710 tok/s).

Inference providers

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

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB7406 tok/scomfortable
GeForce RTX 509032GB11107 tok/scomfortable
M4 Max 128GB128GB2710 tok/scomfortable
M4 Pro 48GB48GB1355 tok/scomfortable
M4 Pro 24GB24GB1355 tok/scomfortable
A100 PCIe 80 GB80GB13570 tok/scomfortable
H100 SXM5 80 GB80GB27293 tok/scomfortable
GeForce RTX 309024GB6548 tok/scomfortable
Radeon RX 7900 XTX24GB5415 tok/scomfortable
GeForce RTX 408016GB5256 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0