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

GLM-4.6V-FP8

by zai-org

108B params · image-text-to-text · 31 likes · 6.0k downloads

GLM-4.6V-FP8 is a 108B parameter model. At Q4 quantization it requires 54GB of VRAM. It runs comfortably on H100 SXM5 80 GB (44 tok/s), H200 SXM 141 GB (65 tok/s).

Inference providers

Provider$/1M in$/1M outThroughput
Z.ai54 tok/s

GPU compatibility

GPUVRAMQ4 DecodeVerdict
M4 Max 128GB128GB4 tok/stight
A100 PCIe 80 GB80GB22 tok/stight
H100 SXM5 80 GB80GB44 tok/scomfortable
M4 Max 64GB64GB4 tok/stight
H200 SXM 141 GB141GB65 tok/scomfortable
M2 Ultra 192GB192GB6 tok/stight
Install CLI [email protected] Raw data · MIT · API data: live · HW/Cloud data: curated 2026-02-23 · v0.6.0