Grux AI Download

Can I run GLM locally on my Mac?

GLM-4.7-Flash runs locally; this site rates it on no Mac. Install Ollama and run ollama run glm-4.7-flash, a 30B-A3B mixture-of-experts model that Ollama lists as a 19GB download with a 198K context window; Ollama's page says it requires Ollama 0.14.3, and the model's card gives the licence as MIT. The 8-bit and bf16 downloads are 32GB and 60GB. GLM-5.2, GLM-5.3 and GLM-5.3-flash are different: Ollama lists them only as cloud models, such as glm-5.3:cloud, so they do not run locally and what you send goes to Ollama's cloud. GLM is not in Grux's catalogue, so this site rates no GLM model on any Mac and quotes Ollama's download sizes only, which are usually smaller than the memory a model needs while running.

Download Grux AI 3.3.1, free, for macOS 14 or later on Apple silicon.

Grux AI 3.3.1 · sources read 2026-11-19 · published 2026-11-19

The GLM models on Ollama

ModelCommandOllama listsWhere it runs
GLM-4.7-Flash, 30B-A3Bollama run glm-4.7-flash19GB (q4_K_M), 198K contextLocally
GLM-4.7-Flash, 8-bitollama run glm-4.7-flash:q8_032GBLocally
GLM-4.7-Flash, bf16ollama run glm-4.7-flash:bf1660GBLocally
GLM-5.2ollama run glm-5.2:cloudCloud tag onlyOllama's cloud
GLM-5.3ollama run glm-5.3:cloudCloud tag only, 1M contextOllama's cloud
GLM-5.3-flashollama run glm-5.3-flash:cloudCloud tag only, 1M context, text and imageOllama's cloud

What is known about GLM-4.7-Flash

GLM-5 on Ollama is a cloud model

Ollama's pages for GLM-5.2, GLM-5.3 and GLM-5.3-flash list a cloud tag and nothing else. The GLM-5.3 and GLM-5.3-flash pages say the model is hosted in the United States and Europe under Ollama's privacy policy, which they describe as zero data retention; the GLM-5.2 page does not say this. Prompts sent to a cloud tag leave your Mac. Ollama's FAQ describes how to turn its cloud features off. This page does not cover running GLM-5 weights by any other route.

What this site does not say

Grux's catalogue lists no GLM model, so the Mac picker does not rate one and this page writes no rating for it on any memory size. Nothing on this page was run on a Mac. For models that are rated, see Ollama models.

Where to check this

ClaimSource
GLM-4.7-Flash: 19GB, 32GB and 60GB, 198K context, Ollama 0.14.3Ollama library, glm-4.7-flash
GLM-4.7-Flash is a 30B-A3B model under the MIT licenceHugging Face, zai-org/GLM-4.7-Flash
GLM-5.3 and GLM-5.3-flash: cloud tag only, hosted in the United States and EuropeOllama library, glm-5.3 and glm-5.3-flash
GLM-5.2: cloud tag onlyOllama library, glm-5.2

Questions

Can I run GLM locally on a Mac?
GLM-4.7-Flash runs locally: ollama run glm-4.7-flash, a 19GB download, so only on a Mac with more memory than that. Ollama lists GLM-5.2 and GLM-5.3 only as cloud models, which do not run locally.
How big is GLM-4.7-Flash?
Ollama lists 19GB for the default q4_K_M, 32GB for q8_0 and 60GB for bf16. Grux's catalogue does not include it, so this site gives no rating for any Mac.
Is GLM-4.7-Flash free for commercial use?
Its Hugging Face card gives the licence as MIT.
Does a GLM cloud tag keep my prompts private?
A cloud tag sends your prompts to Ollama's servers. Ollama's GLM-5.3 and GLM-5.3-flash pages say they are hosted in the United States and Europe with zero data retention, and the GLM-5.2 page does not say where it is hosted; read Ollama's privacy policy for the terms.
Download Grux AI 3.3.1 Read the source

Free, MIT licensed. macOS 14 or later, Apple silicon. 29.1 MB, signed and notarized by Apple. No account, no server, no subscription.