RUN THIS LLM
Search local LLM hardware requirements
GLM-5.1 754B MoE
Zhipu · 754B MoE · General
Frontier 754B MoE with 40B active params. MLA and DeepSeek Sparse Attention. Next-gen flagship for agentic engineering, coding, and long-horizon tool use.
VRAM Requirements
| Quantization | VRAM |
|---|---|
| Q4_K_M (smallest) | 452.4 GB |
| Q8_0 (balanced) | 829.4 GB |
| FP16 (full quality) | 1508 GB |
Specifications
- Parameters: 754B MoE (40B active per token)
- Category: General
- Max context: 200K tokens
- System RAM: 480 GB minimum
- HuggingFace: zai-org/GLM-5.1
Benchmarks
- MATH: 95 — Competition-level math reasoning
Loading interactive analysis...