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

Nemotron-Orchestrator-8B

by nvidia

16.4B params · text-generation · 561 likes · 11.4k downloads

Nemotron-Orchestrator-8B is a 16.4B parameter model. At Q4 quantization it requires 8GB of VRAM. It runs comfortably on GeForce RTX 4090 (80 tok/s), GeForce RTX 5090 (120 tok/s), A100 PCIe 80 GB (146 tok/s).

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB80 tok/scomfortable
GeForce RTX 509032GB120 tok/scomfortable
M4 Max 128GB128GB29 tok/stight
M4 Pro 48GB48GB14 tok/stight
M4 Pro 24GB24GB14 tok/stight
A100 PCIe 80 GB80GB146 tok/scomfortable
H100 SXM5 80 GB80GB295 tok/scomfortable
GeForce RTX 309024GB70 tok/scomfortable
Radeon RX 7900 XTX24GB58 tok/scomfortable
GeForce RTX 408016GB56 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0