Back to Calculator Deploy on RunPodDeploy Now
GLM-5
HotZ.ai's frontier MoE model (745B total, 44B active), trained entirely on Huawei Ascend chips. SOTA open-source SWE-bench performance (77.8%), focuses on agentic engineering and coding. Uses DeepSeek Sparse Attention.
Specifications
SourceArchitectureTEXT
Parameters745B
Familyglm
VRAM (Q4)372.5G
MoE: 44B active.
flagshipzhipumoetrending
Run in the Cloud
This model requires enterprise-grade VRAM. Rent GPUs on RunPod and start generating.
Instant Cloud GPUs
Running out of VRAM? Rent a high-end H100 or RTX 4090 on RunPod and deploy in seconds.
Quantization Estimates
| Format | VRAM Need | Tier |
|---|---|---|
| FP16 | 1490.0 GB | Full Precision |
| Q8_0 | 745.0 GB | High |
| Q6_K | 633.3 GB | Excellent |
| Q5_K_M | 521.5 GB | Great |
| Q4_K_M | 372.5 GB | Sweet Spot |
| Q2_K | 223.5 GB | Emergency |
Share this Model
Send these specs directly to your community.