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

distilbert-base-multilingual-cased-ner-hrl

by Davlan

135M params · token-classification · 84 likes · 64.1k downloads

distilbert-base-multilingual-cased-ner-hrl is a 135M parameter model. At Q4 quantization it requires 0GB of VRAM. It runs comfortably on GeForce RTX 4090 (9744 tok/s), GeForce RTX 5090 (14613 tok/s), M4 Max 128GB (3565 tok/s).

Inference providers

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

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB9744 tok/scomfortable
GeForce RTX 509032GB14613 tok/scomfortable
M4 Max 128GB128GB3565 tok/scomfortable
M4 Pro 48GB48GB1782 tok/scomfortable
M4 Pro 24GB24GB1782 tok/scomfortable
A100 PCIe 80 GB80GB17853 tok/scomfortable
H100 SXM5 80 GB80GB35908 tok/scomfortable
GeForce RTX 309024GB8615 tok/scomfortable
Radeon RX 7900 XTX24GB7124 tok/scomfortable
GeForce RTX 408016GB6915 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0