Autor: Jakub Rusinowski · Ostatnia aktualizacja: 1 maja 2026
Model library → GLM-5 / GLM-5.1 → GLM-5.1 72B
GLM-5.1 is the refined successor to GLM-5 with improved instruction following, reduced hallucinations, and enhanced agentic workflow support. At 72B parameters with MIT license, it's among the most capable fully open models available for commercial use.
GLM-5.1 72B needs about 44 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.
| Parameters | 72 Billion |
| Context window | 128,000 |
| Architecture | GLM (General Language Model) |
| Provider | Zhipu AI (Z.ai) |
| Licence | MIT |
| Specified at | Q4_K_M |
| System RAM | 64 GB |
| Record updated | 2026-05-01 |
MIT — commercial use permitted. Commercial use permitted. No usage restrictions beyond attribution.
Modelled on a reference NVIDIA RTX 4090 (24 GB), with no KV cache (this record has no published architecture). Speed figures are ESTIMATES from the memory-bandwidth roofline described on the methodology page, not benchmarks we ran — rows marked measured come from published or reader-submitted runs. VRAM here includes the KV cache, so it reads higher than the headline figure above, which does not.
| Quant | Weights | VRAM needed | Est. speed | Fit on 24 GB |
|---|---|---|---|---|
| Q2_K | 23.7 GB | 24.5 GB | ~4 tok/s (est.) | Offloads to system RAM (slow) |
| Q3_K_M | 30.7 GB | 31.5 GB | ~3 tok/s (est.) | Offloads to system RAM (slow) |
| Q4_K_M | 43.5 GB | 44.3 GB | ~2 tok/s (est.) | Offloads to system RAM (slow) |
| Q5_K_M | 51.0 GB | 51.8 GB | ~2 tok/s (est.) | Offloads to system RAM (slow) |
| Q6_K | 59.0 GB | 59.8 GB | — | Won't fit |
| Q8_0 | 76.5 GB | 77.3 GB | — | Won't fit |
| F16 | 144.0 GB | 144.8 GB | — | Won't fit |
Want the memory numbers alone, at every quantization level and your own context length? Use the GLM-5.1 72B VRAM calculator.
As an Amazon Associate we earn from qualifying purchases. Cloud GPU links are referral links — we may earn a commission at no extra cost to you.
The cheapest catalogued GPU that runs GLM-5.1 72B is the Apple M5 Pro (64 GB).
Install Ollama, then run:
ollama run hf.co/THUDM/GLM-5.1-72B-Chat-Q4_K_MWeights on Hugging Face: THUDM/GLM-5.1-72B-Chat.
Quality scores as published by the model's authors or an independent evaluator — not throughput, and not measured by us.
| Benchmark | Score | Provenance |
|---|---|---|
| MMLU | 89.1 / 100 % | reported |
| HumanEval | 92.6 / 100 % | reported |
| AgentBench | 74.3 / 100 % | reported |
Best for: agentic, reasoning, enterprise, complex tasks, long context.
← All GLM-5 / GLM-5.1 models | VRAM calculator | Check your own hardware