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

starcoder2-15b

by bigcode

16.0B params · text-generation · 673 likes · 10.9k downloads

starcoder2-15b is a 16.0B parameter model. At Q4 quantization it requires 8GB of VRAM. It runs comfortably on GeForce RTX 4090 (82 tok/s), GeForce RTX 5090 (123 tok/s), M4 Max 128GB (30 tok/s).

GPU compatibility

GPUVRAMQ4 DecodeVerdict
GeForce RTX 409024GB82 tok/scomfortable
GeForce RTX 509032GB123 tok/scomfortable
M4 Max 128GB128GB30 tok/scomfortable
M4 Pro 48GB48GB15 tok/stight
M4 Pro 24GB24GB15 tok/stight
A100 PCIe 80 GB80GB150 tok/scomfortable
H100 SXM5 80 GB80GB303 tok/scomfortable
GeForce RTX 309024GB72 tok/scomfortable
Radeon RX 7900 XTX24GB60 tok/scomfortable
GeForce RTX 408016GB58 tok/scomfortable
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0