Question 1

How much VRAM does GLM-4.7 need?

Accepted Answer

GLM-4.7 (358.3B) requires approximately 800.9473784923554 GB of VRAM at full precision (FP16), 400.5 GB at 8-bit quantization (Q8), and 200.2 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-4.7?

Accepted Answer

We currently have no compatible GPUs indexed for GLM-4.7. This may be because the model's VRAM requirements exceed available GPUs in our database, or VRAM data hasn't been computed yet.

Question 3

What is the cheapest GPU to run GLM-4.7?

Accepted Answer

We don't currently have cloud pricing data for GPUs compatible with GLM-4.7. Cloud pricing data is updated regularly — check back soon.

Question 4

Can GLM-4.7 run on consumer hardware?

Accepted Answer

GLM-4.7 requires approximately 800.9473784923554 GB of VRAM at FP16, and even with 4-bit quantization (Q4) it needs about 200.2 GB. This exceeds the 24 GB available on the highest-end consumer GPU (RTX 4090), so GLM-4.7 requires datacenter or cloud GPUs.

Question 5

What quantization options are available for GLM-4.7?

Accepted Answer

GLM-4.7 supports the following quantization levels: FP16 (full precision): ~800.9473784923554 GB, Q8 (8-bit): ~400.5 GB, Q6 (6-bit): ~300.4 GB, Q4 (4-bit): ~200.2 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. As a Mixture of Experts (MoE) model with 160 experts (8 active), GLM-4.7 has a larger total weight size but only activates a subset of parameters per token, which can affect effective memory usage during inference.

GLM-4.7

VRAM Requirements

Compatible GPUs

GLM-4.7 FAQ

VRAM requirements for GLM-4.7 at different quantization levels
Quantization	VRAM Required
FP16	800.9 GB
Q8	400.5 GB
Q6	300.4 GB
Q4	200.2 GB