How much memory does GLM-5.3 need?
About 462 GB at Q4_K_M with an 8K context. That is more than any single Mac.
Try other settings in the calculatorMemory by quantization and context
Total memory: weights, KV cache in F16 and about 1 GB of buffers. Under each total, the smallest Mac that can run it with the default GPU memory limit.
| Quantization | 8K context | 32K context | 128K context |
|---|---|---|---|
| Q3_K_M | 370 GBtoo big for a Mac | 372 GBtoo big for a Mac | 381 GBtoo big for a Mac |
| Q4_K_M | 462 GBtoo big for a Mac | 464 GBtoo big for a Mac | 473 GBtoo big for a Mac |
| Q5_K_M | 538 GBtoo big for a Mac | 541 GBtoo big for a Mac | 550 GBtoo big for a Mac |
| Q6_K | 619 GBtoo big for a Mac | 622 GBtoo big for a Mac | 631 GBtoo big for a Mac |
| Q8_0 | 802 GBtoo big for a Mac | 804 GBtoo big for a Mac | 813 GBtoo big for a Mac |
| F16 | 1,508 GBtoo big for a Mac | 1,511 GBtoo big for a Mac | 1,519 GBtoo big for a Mac |
About the model
- Parameters
- 753B
- Type
- Mixture of experts, 256 experts
- Attention
- compressed latent attention (MLA)
- KV cache, F16
- 0.7 GB at 8K, 12 GB at 128K
- On Hugging Face
- zai-org/GLM-5.3
Each layer caches one compressed latent per token instead of full keys and values, which keeps the KV cache small for a model this size. As a mixture-of-experts model it loads every expert into memory, so memory follows the total parameter count, while speed follows the smaller number of active parameters.
Questions
How much memory does GLM-5.3 need?
About 462 GB at Q4_K_M with an 8K context: 460 GB of weights, 0.7 GB of KV cache and about 1 GB of buffers. At 32K context it needs about 464 GB.
Can GLM-5.3 run on a Mac?
Not on a single Mac at Q4_K_M: it needs about 462 GB, more than the largest Mac memory size. It needs a cluster of machines or a lower-bit quantization.
How big is the GLM-5.3 KV cache?
About 0.7 GB at 8K tokens and 12 GB at 128K tokens with an F16 cache. Its attention design: compressed latent attention (MLA).
Other models
- Gemma 4 12B
- gpt-oss 20B
- Gemma 4 26B-A4B
- Qwen3.8 27B
- GLM-4.7 Flash
- Gemma 4 31B
- Qwen3.6 35B-A3B
- gpt-oss 120B
- Mistral Medium 3.5
- Qwen3.8 Flash Next
- DeepSeek V4 Flash
- DeepSeek V4 Pro
- Kimi K3
Data read from Hugging Face on 2026-11-09. These are estimates; benchmarks on this site report measured memory. How the estimate works. Compare all models in the model fit chart.
Get the next one
New benchmarks and agent runs by email, with the raw data. Or follow on RSS or X.