Can You Run GLM-4.7-Flash (30B-A3B)? VRAM Requirements
GLM-4.7-Flash (30B-A3B) has 31.2B total parameters (MoE, 3B active per token). Running it at Q4_K_M (GGUF) with 16K context takes roughly 20.1 GB of memory (17.7 GB weights + 0.8 GB KV cache + runtime overhead). The minimum viable hardware is about RTX 3090 / 4090 24GB. All numbers are planning estimates, verified 2026-09-05.
VRAM by quantization (16K context)
| Quantization | Weights | Total needed | Minimum hardware |
|---|---|---|---|
| FP16 / BF16 | 58.1 GB | 63.6 GB | A100 / H100 80GB |
| FP8 | 29.1 GB | 32.2 GB | Mac 36GB unified |
| Q8_0 (GGUF) | 30.8 GB | 34.1 GB | A100 40GB |
| Q6_K (GGUF) | 23.8 GB | 26.6 GB | RTX 5090 32GB |
| Q5_K_M (GGUF) | 20.6 GB | 23.1 GB | RTX 5090 32GB |
| MXFP4 | 15.4 GB | 17.7 GB | RTX 3090 / 4090 24GB |
| NVFP4 | 15.4 GB | 17.7 GB | RTX 3090 / 4090 24GB |
| Q4_K_M (GGUF)recommended | 17.7 GB | 20.1 GB | RTX 3090 / 4090 24GB |
| Q3_K_M (GGUF) | 13.7 GB | 16.0 GB | RTX 3090 / 4090 24GB |
| Q2_K (GGUF) | 10.2 GB | 12.5 GB | RTX 4060 Ti 16GB |
How context length changes memory (Q4_K_M (GGUF))
| Context | KV cache | Total needed |
|---|---|---|
| 4K | 0.2 GB | 19.4 GB |
| 16K | 0.8 GB | 20.1 GB |
| 64K | 3.3 GB | 22.5 GB |
| 128K | 6.6 GB | 25.8 GB |
KV cache grows linearly with context; FP8 KV cache halves it again (try it in the calculator below).
Customize the estimate
LLM VRAM Calculator — Can I Run It?
Estimate how much GPU memory (VRAM) you need to run open-weight LLMs like Kimi K3, DeepSeek R1, Qwen3, or Llama locally. Pick a model, quantization, and context length — the calculator adds up weights, KV cache, and runtime overhead, then shows which GPUs or Macs can fit it. Runs entirely in your browser: no upload, no signup.
31.2B total params · 3B active (MoE) · max 198K context
Recommended balance · ~0.61 bytes/param
Estimated memory needed
25.8 GB
Will it fit?
Assumes ~92% of device memory is usable. Multi-GPU counts are for tensor/pipeline parallel serving (vLLM, SGLang); Apple Silicon uses unified memory via llama.cpp or MLX.
FAQ
How much VRAM does GLM-4.7-Flash (30B-A3B) need?
About 20.1 GB at the recommended Q4_K_M (GGUF) quantization with 16K context: 17.7 GB for weights, 0.8 GB for KV cache, plus runtime overhead. Longer context grows the KV cache.
What is the minimum hardware for GLM-4.7-Flash (30B-A3B)?
Roughly RTX 3090 / 4090 24GB, assuming 92% of device memory is usable. Lower-end setups can try more aggressive quantization (Q3/Q2) at a noticeable quality cost.
Can I run GLM-4.7-Flash (30B-A3B) on a Mac?
Yes. At Q4_K_M (GGUF) with 16K context it needs about 20.1 GB, so a Mac 24GB unified or larger fits it via llama.cpp or MLX. Note macOS only lets the GPU wire roughly 2/3-3/4 of RAM by default; near the limit, raise iogpu.wired_limit_mb or use a smaller quant.
Which quantization should I use for GLM-4.7-Flash (30B-A3B)?
Q4_K_M is the recommended balance of size and quality; with headroom, Q6_K or Q8_0 reduce quality loss further.