Skip to content

Text model · GLM

GL GLM-4.6: RAM and VRAM requirements

GLM-4.6 needs about 221.3 GB to run at Q4_K_M (the Q4_K_M GGUF file is ~216 GB to download; KV cache and overhead add the rest), or about 384.3 GB at Q8_0. The lightest hardware that runs it is a high-memory machine.

GLM family · 357B params (Mixture-of-Experts: activates only 32B of 357B params per token, so generation is faster than the total size suggests) · released Aug 2025.

Shopping for hardware? See what runs GLM-4.6 →

License MIT · Commercial OK ↓ 26.5K/mo ♥ 1.2K on HuggingFace
Q4_K_M GGUF
216 GB
Q8_0 GGUF
379 GB
Memory @ Q4 (4k)
~221.3 GB
Context
200 k

Q4_K_M and Q8_0 are GGUF download (file) sizes. Memory @ Q4 includes the KV cache and overhead, so it runs larger than the file on disk.

Will it run on your device?

GLM-4.6 runs on 0 of 40 tracked devices at Q4_K_M.

0 run well 0 tight fit 40 too small
Fit check Q4_K_M
No, not enough memory
needs 221.3 GB usable 10.5 GB

Memory breakdown

Weights (Q4_K_M)216 GB
+
KV cache (4k)4.5 GB
+
Overhead0.8 GB
=
Total221.3 GB

How context length changes it

4k context ~221.3 GB 32k context ~253.1 GB 128k context ~361.9 GB

Longer context grows the KV cache, which for GLM-4.6 is sized by its 32B active params, not the full 357B. It needs ~221.3 GB at 4k and ~361.9 GB at 128k.

Quantization sizes

GGUF quantson disk
Quantization Size on disk
Q2_K 149.5 GB est
Q3_K_M 174.5 GB est
Q4_K_M (default) 216 GB
Q5_K_M 254.4 GB est
Q6_K 292.7 GB est
Q8_0 379 GB
FP16 714 GB

Lower quant = smaller and faster, slightly lower quality. Q4_K_M is the common default.

Run it

llama.cpp
$ llama-cli -hf unsloth/GLM-4.6-GGUF:Q4_K_M
LM Studio
$ lms get unsloth/GLM-4.6-GGUF

Which devices can run GLM-4.6?

Same job, different size

Pick by what fits your memory: step down to free up VRAM, or step up if you have headroom.

Similar models

Head-to-head

FAQ

How much VRAM or RAM does GLM-4.6 need?

At Q4_K_M, GLM-4.6 needs about 221.3 GB (weights ~216 GB + KV cache + overhead) at a 4k context. At Q8_0 budget ~384.3 GB.

What is the Q4_K_M GGUF file size of GLM-4.6?

The Q4_K_M GGUF file is about 216 GB to download, and the Q8_0 GGUF is about 379 GB. That is the weights file on disk; to run it you also need room for the KV cache and overhead, so budget ~221.3 GB of memory at Q4_K_M.

Can GLM-4.6 run on a laptop?

GLM-4.6 is large; you need a high-memory Mac or multi-GPU setup at Q4_K_M.

Is GLM-4.6 cheaper to run because it is a MoE model?

It is faster, not lighter. GLM-4.6 activates only 32B of 357B params per token (so it runs quickly), but all experts must stay in memory, so it still needs memory for the full 357B.

Can I use GLM-4.6 commercially?

Yes. GLM-4.6 is licensed MIT, which permits commercial use.

Understand the numbers

Short guides to the ideas behind GLM-4.6's memory and quant figures.

Z.ai's GLM-4.6 flagship: 357B MoE, 32B active, 200K context, MIT. Needs a multi-GPU rig or a large Mac cluster.

Sources

Last validated 2026-08-03. Memory figures are estimates. See methodology.