Support Matrix#
Hardware#
Unless specified otherwise, NVIDIA NIM for vision language models (VLMs) should, but are not guaranteed to, run on any NVIDIA GPU, provided the GPU has sufficient memory. They can also run on multiple homogeneous NVIDIA GPUs with sufficient aggregate memory and a CUDA compute capability of >= 7.0 (8.0 for bfloat16) unless otherwise specified. For more information, refer to Supported Models.
NVIDIA NIM for VLMs does not support NVIDIA Virtual GPU (vGPU) environments.
For information on the supported operating systems, drivers, and software, refer to the About Get Started page.
Supported Models#
Kimi-K2.6#
Latest supported release tag: 2.0.9-variant
The following section lists the supported configurations for
moonshotai/kimi-k2.6-turbo
(NGC catalog page).
Generic Configuration#
Kimi-K2.6 is a ~1T-parameter MoE VLM. It ships in two precision lanes selected automatically by GPU architecture: INT4 on Hopper (TP8) and NVFP4 on Blackwell (TP4).
The GPU Memory column is per-GPU HBM in GB; the Disk Space column is the NGC artifact size needed in the NIM cache (one-time download on first launch), in GB.
GPU |
GPU Memory |
Precision |
# of GPUs |
Disk Space |
|---|---|---|---|---|
NVIDIA-H200 |
141 |
INT4 |
8 |
595 |
NVIDIA-B200 |
192 |
NVFP4 |
4 |
595 |
NVIDIA-B300-SXM6-AC |
288 |
NVFP4 |
4 |
595 |
NVIDIA-GB200 |
192 |
NVFP4 |
4 |
595 |
NVIDIA-GB300 |
288 |
NVFP4 |
4 |
595 |
Muse Glimmer#
Latest supported release tag: 2.1.3-variant
The following section lists the supported configurations for
meta/muse-glimmer
(NGC catalog page).
Generic Configuration#
The GPU Memory and Disk Space values are in GB; Disk Space is for both the container and the model. Unless indicated, the first number indicates the disk space required without DFlash, and the second number indicates the disk space required with DFlash.
Any denotes the fallback profile.
GPU |
GPU Memory |
Precision |
# of GPUs |
Disk Space |
|---|---|---|---|---|
Any |
> 80 |
BF16 |
1 |
60, 65 |
NVIDIA-B200 |
196 |
BF16 |
1 |
60, 65 |
NVIDIA-H200 |
141 |
BF16 |
1 |
60, 65 |
NVIDIA-H100 |
80 |
BF16 |
1 |
60, 65 |
NVIDIA-H20 |
96 |
BF16 |
1 |
60, 65 |
NVIDIA-H20-3e |
141 |
BF16 |
1 |
60, 65 |
NVIDIA-DGX-Spark |
128 |
BF16 |
1 |
65 with DFlash |