Can you run GLM-5.3 locally?
zai-org/GLM-5.3 · 753B parameters · license glm-5.3 · on Hugging Face since 2026-08-25
Short answer: not on one tracked machine
The weights alone need about 457 GB at Q4_K_M and 800 GB at Q8_0 (estimates from parameter count × bits per weight). The largest tracked machine has 256GB, so no single box here can hold it at 4-bit.
What it would take (estimate)
Assumes 4-bit weights × 1.15 + 2 GB ≈ 527 GB, spread over machines with 85% of their memory usable. Clustering needs software such as llama.cpp RPC and fast networking; we have no public speed result for this model on any tracked setup.
| Option | Hardware cost | EUR |
|---|---|---|
| 5 × Framework Desktop 128GB (DIY) (128GB each) | ≈$17,245 | ≈€19,445 |
| 3 × Mac Studio M5 Ultra 80-core GPU, 256GB / 1TB (256GB each) | ≈$32,397 | ≈€37,587 |
Or rent it
| Cloud GPUs (80GB each) | USD | EUR |
|---|---|---|
| 7 × H100 80GB, RunPod (Community Cloud (Secure $2.89)) | ≈$13.93/hour | ≈€12.38/hour |
| 7 × H100 80GB, Vast.ai (median of 2 live offers) | ≈$21.72/hour | ≈€19.30/hour |
Hourly cost = GPUs needed × tracked H100 PCIe hourly price (see Rent instead). Multi-GPU machines with that many cards may not be offered at this price.
Smaller siblings that fit
GLM-5.3-Flash (≈194.8 GB at 4-bit)
Why it's listed: 1.8M downloads; same family as GLM-5.3-Flash.