GLM 4.7 Flash

Overview

Zhipu AI's compact 3B variant of GLM 4.7, MIT-licensed with a 128k context. Optimized for low-latency bilingual Chinese-English chat.

When to pick this model

• Bilingual zh/en chat assistants where latency is critical
• Lightweight chat backends with a strict permissive license requirement
• Long-context summarization on small GPUs
• Cost-sensitive serving at scale where 30B variants are overkill

VRAM requirements by quantization

Quantization	VRAM required
Q4_K_M (recommended)	1.7 GB
Q5_K_M	2.1 GB
Q8_0	3.2 GB
FP16 (no quantization)	6 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

Strengths

• MIT license — among the most permissive in the open ecosystem
• 128k context in a 3B footprint
• Strong Chinese and English performance
• Compact ~1.7GB VRAM at Q4

Limitations

• Gated on Hugging Face despite the open license
• Less versatile than the 30B GLM 4.7 variants

Architecture & training

Architecture: Dense transformer · 3B parameters · 128k context

Training: GLM 4.7 family from Zhipu AI / THUDM (Tsinghua). Flash variant optimized for latency, focus on zh/en.

Quick start

ollama run glm-4.7-flash

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.