What GPU do I need to run zai-org/GLM-4.7-Flash?
A 31.2B (MoE) language model for chat and instruction-following. 31.2B parameters, published in BF16. View on Hugging Face
GLM-4.7-Flash is published by zai-org on Hugging Face, with 1,948,824 downloads and 1,831 likes to date. It's a Glm4MoeLiteForCausalLM model built for text-generation, published natively in BF16.
What GLM-4.7-Flash is
GLM-4.7-Flash is a 31.2B-parameter mixture-of-experts language model published by Z.ai (Zhipu) on Hugging Face. It is released under MIT.
License note: permissive: allows commercial use, modification and redistribution. Facts in this section are sourced from GLM-4.7-Flash's Hugging Face model card, not benchmarked by Aquanode.
What it's used for
- Chat assistants
- Instruction following
- Synthetic data generation
VRAM required & cheapest live GPU fit
Required VRAM = weight size at each precision times a flat 1.2 overhead for activations and fragmentation. The KV-cache grows with context and is not in that factor; it is listed per model below where the architecture is published. Full formula and assumptions: methodology.
| Precision | Weight size | Required VRAM | Cheapest live fit | GPUs needed | Est. $/hr (full fit) |
|---|---|---|---|---|---|
| BF16 | 58.2 GB | 69.8 GB | A100 | 1 | $1.18/hr |
| cheaper alt. | RTX A5000 | 3 | $0.528/hr | ||
| FP8 (quantized) | 29.1 GB | 34.9 GB | L40 | 1 | $0.759/hr |
| cheaper alt. | RTX 5060 Ti | 3 | $0.435/hr | ||
| INT4 (quantized) | 14.5 GB | 17.4 GB | RTX A5000 | 1 | $0.176/hr |
A GPU is only matched to a row if its hardware supports that precision, and the primary recommendation is always a single-GPU fit when one exists.
INT4 requires a quantized checkpoint actually published for this model, check its Hugging Face page before relying on this row.
Cheapest way to run GLM-4.7-Flash at its published (BF16) precision: 1× A100, at $1.18/hr per GPU ($1.18/hr total). Quantizing to FP8 or INT4 (rows above) can cost less, but requires a compatible quantized checkpoint to exist for this model.
GLM-4.7-Flash: KV cache by context length
The KV cache is the memory the attention layers hold for every token of context, on top of the weights. It is not part of the flat 1.2x overhead in the table above, grows with context length and with every concurrent request, and is why a long-context deployment needs more VRAM than the table shows.
Multi-head latent attention: every layer caches one compressed latent per token. Cached per token: 54,144 bytes at 16-bit.
| Context | KV cache, one sequence |
|---|---|
| 4K | 0.21 GB |
| 32K | 1.7 GB |
| 128K | 6.6 GB |
| 198K (model maximum) | 10.2 GB |
Computed from the layer, head and window counts in the model's own configuration (zai-org/GLM-4.7-Flash), at 16 bits per cached value. Each concurrent sequence needs its own cache, on top of the weights.
GLM-4.7-Flash: common questions
Can GLM-4.7-Flash run on a single GPU?
Yes, but not on a desktop card. At BF16 it needs 69.8 GB of VRAM, more than a single 32 GB desktop card holds. The cheapest live single-card fit is a 80.0 GB A100 at $1.18/hr.
What is the least VRAM GLM-4.7-Flash can run in?
17.4 GB, at INT4 (quantized), which fits a 24 GB card, against 69.8 GB at BF16. That row assumes an INT4 checkpoint has actually been published for this model, so check its Hugging Face page before planning around it.
Does quantizing GLM-4.7-Flash lower the GPU bill?
Yes. At BF16 the cheapest live fit is one A100 at $1.18/hr. At INT4 (quantized) it drops to one RTX A5000 at $0.176/hr, provided a quantized checkpoint exists for it.
How to run GLM-4.7-Flash
Run GLM-4.7-Flash with vLLM
From zai-org/GLM-4.7-Flash's own deployment docs.
vllm serve zai-org/GLM-4.7-Flash \
--tensor-parallel-size 4 \
--speculative-config.method mtp \
--speculative-config.num_speculative_tokens 1 \
--tool-call-parser glm47 \
--reasoning-parser glm45 \
--enable-auto-tool-choice \
--served-model-name glm-4.7-flashSource: https://huggingface.co/zai-org/GLM-4.7-Flash/raw/main/README.md
Run GLM-4.7-Flash with Ollama
Verified against Ollama's own library listing.
ollama run glm-4.7-flashRun GLM-4.7-Flash with GGUF quantizations
Prebuilt GGUF weights published at unsloth/GLM-4.7-Flash-GGUF. Run with llama.cpp's llama-server or load the repo directly in LM Studio.
llama-server -hf unsloth/GLM-4.7-Flash-GGUFDeploy GLM-4.7-Flash on Aquanode
Aquanode has no one-click deploy template for GLM-4.7-Flash; you install the inference engine yourself with the commands below. Aquanode sells GPU pods billed per second, not a hosted inference API.
- Launch a bare GPU pod sized to the requirement above (1× A100 or larger).
- Open a terminal on the pod, or save one of the commands above as a startup script so it runs automatically the first time the pod boots.
- Run the command and connect to the resulting endpoint.
Weight-to-VRAM math, the fit rules, and how live prices are normalized: full methodology.
More GLM-4.5 models
- GLM-4.6V-Flash (10.3B, BF16)
- GLM-4.5V (107.7B, BF16)
- GLM-4.5-Air (110.5B, BF16)
- GLM-4.5-Air-Base (110.5B, BF16)
- GLM-4.5-Air-FP8 (110.5B, F8_E4M3)
Lineage
Also published as: unsloth/GLM-4.7-Flash. Identical copies of this model under another account.
Alternatives at this size
Other models for text-generation within about a third of GLM-4.7-Flash's 31.2B parameters, from other model lines.
- Qwen3-32B (32.8B, BF16)
- NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 (31.6B, BF16)
- granite-4.1-30b (28.9B, BF16)
- Ornith-1.5-35B-A3B (36.0B, BF16)
- Phi-3.5-MoE-instruct (41.9B, BF16)
More on GLM-4.7-Flash
Fits on an 80 GB GPU at BF16: every model that fits in 80 GB.
Fits on a 48 GB GPU at FP8: every model that fits in 48 GB.
Fits on a 24 GB GPU at INT4: every model that fits in 24 GB.
Best chat and assistants models: how GLM-4.7-Flash ranks against the rest.
Compare GLM-4.7-Flash with Qwen3-32B, Qwen3-30B-A3B, Qwen2.5-32B-Instruct, Qwen2.5-Coder-32B-Instruct, Qwen3-30B-A3B-Instruct-2507, NVIDIA-Nemotron-3-Nano-30B-A3B-BF16, Qwen3-Coder-30B-A3B-Instruct.
Related reading: A100 pricing and specs, How much VRAM you need for LLMs, Serving LLMs with vLLM, vLLM vs TensorRT-LLM vs SGLang, and Best GPU for LLM inference.
Related on Aquanode: GPU requirements for Falcon3-1B-Instruct, GPU requirements for Kolibri-1, GPU requirements for internlm2-chat-20b, and GPU requirements for llama-3-typhoon-v1.5-8b-instruct.
Your own GPU cloud accounts
One console for the GPU clouds you already rent from. Your keys stay on your machine.
$20 a month, flat