Question 1

How much VRAM does GLM-4.6V need?

Accepted Answer

GLM-4.6V (107.7B) requires approximately 240.7282590866089 GB of VRAM at full precision (FP16), 120.4 GB at 8-bit quantization (Q8), and 60.2 GB at 4-bit quantization (Q4). The actual VRAM usage may be slightly higher due to KV cache and runtime overhead.

Question 2

What GPUs can run GLM-4.6V?

Accepted Answer

We have 2 compatible GPUs indexed for GLM-4.6V, including MI325X, MI350X. Compatibility is determined by whether the GPU's VRAM can fit the model's memory footprint, potentially using quantization to reduce memory usage.

Question 3

What is the cheapest GPU to run GLM-4.6V?

Accepted Answer

We don't currently have cloud pricing data for GPUs compatible with GLM-4.6V. Cloud pricing data is updated regularly — check back soon.

Question 4

Can GLM-4.6V run on consumer hardware?

Accepted Answer

GLM-4.6V requires approximately 240.7282590866089 GB of VRAM at FP16, and even with 4-bit quantization (Q4) it needs about 60.2 GB. This exceeds the 24 GB available on the highest-end consumer GPU (RTX 4090), so GLM-4.6V requires datacenter or cloud GPUs.

Question 5

What quantization options are available for GLM-4.6V?

Accepted Answer

GLM-4.6V supports the following quantization levels: FP16 (full precision): ~240.7282590866089 GB, Q8 (8-bit): ~120.4 GB, Q6 (6-bit): ~90.3 GB, Q4 (4-bit): ~60.2 GB. Lower quantization reduces VRAM usage at the cost of some quality loss. Q8 offers a good balance between quality and memory savings, while Q4 provides maximum compression.

GLM-4.6V

VRAM Requirements

Compatible GPUs

GLM-4.6V FAQ

VRAM requirements for GLM-4.6V at different quantization levels
Quantization	VRAM Required
FP16	240.7 GB
Q8	120.4 GB
Q6	90.3 GB
Q4	60.2 GB

Top 10 compatible GPUs for GLM-4.6V, sorted by cheapest price
GPU	VRAM	Best Quant	From	Rent
MI325X	256 GB	Q8	—
MI350X	288 GB	FP16	—