DEV.co
Open-Source LLM · nvidia

GLM-5-NVFP4

GLM-5-NVFP4 is a 435B-parameter (40B activated) mixture-of-experts language model quantized to NVIDIA's FP4 format. It is a production-ready quantization of ZAI's GLM-5 base model, optimized for inference on NVIDIA GPUs via vLLM or SGLang. The model supports 200K context length and is released under the MIT license, permitting commercial and non-commercial use. It trades minimal accuracy loss (~0.3-1% on benchmarks) for significant inference speedup and reduced memory footprint compared to FP8 baseline.

Source: HuggingFace — huggingface.co/nvidia/GLM-5-NVFP4
435.2B
Parameters
mit
License (OSI-approved)
Unknown
Context (tokens)
34.2k
Downloads (30d)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Developernvidia
Parameters435.2B
Context windowUnknown
Licensemit — OSI-approved
Modality / tasktext-generation
Gated on HuggingFaceNo
Downloads34.2k
Likes30
Last updated2026-04-10
Sourcenvidia/GLM-5-NVFP4

What GLM-5-NVFP4 is

GLM-5-NVFP4 applies post-training quantization (PTQ) using NVIDIA ModelOpt v0.42.0, quantizing weights and activations of linear layers in MoE transformer blocks to NVFP4 precision. Calibration used cnn_dailymail and Nemotron-Post-Training-Dataset-v2. Evaluated on MMLU Pro, GPQA Diamond, SciCode, IFBench, and HLE benchmarks; results show NVFP4 within 0.3–1% of FP8 baseline. Architecture: auto-regressive transformer with GLM-5 topology and MoE routing. Input: 1D text strings; Output: 1D text sequences. Supported runtimes: vLLM, SGLang. Test hardware: B300. Requires Linux + NVIDIA GPU (Blackwell or compatible).

Quickstart

Run GLM-5-NVFP4 locally

Load the open weights with 🤗 Transformers and generate — the same model, self-hosted.

quickstart.pypython
from transformers import pipelinepipe = pipeline("text-generation", model="nvidia/GLM-5-NVFP4")out = pipe("Explain retrieval-augmented generation in one sentence.",           max_new_tokens=128)print(out[0]["generated_text"])

Swap in vLLM or Ollama for production-grade serving. DEV.co can stand up the inference stack.

Deployment

How you'd run it

A typical self-hosted path — open weights, an inference server, your application.

DEV.co builds each layer — from GPU infrastructure to the application.

Best use cases

AI Agent & Chatbot Systems

GLM-5-NVFP4's tool-calling support (glm47 parser), reasoning parser (glm45), and 200K context make it suitable for autonomous agent frameworks and multi-turn conversational AI systems. Quantization reduces deployment latency and GPU memory vs. full-precision alternatives.

Retrieval-Augmented Generation (RAG)

The 200K context window and support for long document processing via vLLM/SGLang chunked prefill enable efficient RAG pipelines. NVFP4 quantization lowers inference cost per token for high-throughput document retrieval scenarios.

Cost-Optimized Private/Self-Hosted LLM Services

For enterprises deploying LLMs on-premise, GLM-5-NVFP4's 4-bit quantization reduces GPU VRAM demands and inference time while maintaining near-FP8 accuracy. MIT license permits commercial self-hosting without licensing friction.

Running & fine-tuning it

Minimum: NVIDIA GPU with ~100–150 GB VRAM for full model (435B params, 40B activated; NVFP4 ≈ 220–260 GB raw weights). Recommended: Multi-GPU setup (e.g., 8×H100/L40S or 4×B300) for tensor parallelism. Example vLLM config uses --tensor-parallel-size 8. Inference tested on B300. Linux OS required. Estimate: 40B activated params at FP4 ≈ 20 GB active memory per GPU with typical batch sizes; full 435B ≈ 217 GB quantized weights. Requires CUDA 12.1+ and compatible vLLM/SGLang builds.

Not supported or documented. Post-training quantization (PTQ) freezes weights in NVFP4 format; further fine-tuning would require de-quantization or continued training in quantized domain (not standard practice). If fine-tuning is needed, consider downloading the unquantized GLM-5 base model (zai-org/GLM-5) and applying LoRA/QLoRA. NVIDIA ModelOpt may support quantization-aware training (QAT), but no evidence in card. Recommend contacting NVIDIA for quantized adaptation workflows.

When to avoid it — and what to weigh

  • Requires Maximum Accuracy on Specialized Benchmarks — While NVFP4 matches FP8 closely, it shows ~1% drops on SciCode and IFBench. If benchmark-specific 0.5% accuracy improvements are critical, FP8 or full-precision GLM-5 is preferable.
  • Non-NVIDIA Hardware or CPU Inference — Model is optimized and tested only on NVIDIA Blackwell GPUs with vLLM/SGLang. CPU inference or non-NVIDIA accelerators (AMD, Intel) are not officially supported and will be significantly slower.
  • Minimal Fine-tuning or Domain Adaptation Expected — The model card provides no fine-tuning guidance, and post-training quantization complicates weight updates. If your use case requires frequent domain-specific LoRA or QLoRA adaptation, a non-quantized base model may be simpler.
  • Production Deployment Without Toxicity/Bias Mitigation — Model card explicitly states the base model inherits toxic language and societal biases from internet crawl data. Requires guardrails and content moderation testing before production use in customer-facing applications.

License & commercial use

MIT License (permissive OSI-approved open-source license). Permits commercial and non-commercial use, modification, and distribution with attribution. No copyleft obligations. Full license text: https://opensource.org/licenses/MIT.

MIT license explicitly permits commercial use. Model card states: 'This model is ready for commercial/non-commercial use.' No gating, no subscription required. However, note: (1) Model is a derivative of ZAI's GLM-5 base model (separate license review recommended); (2) Base model inherits internet-crawled biases and toxicity—vendors must implement content guardrails and test for use-case compliance; (3) NVIDIA disclaims ownership and responsibility; ZAI's original GLM-5 license should be reviewed for any upstream restrictions. Summary: Commercial use is permissible under MIT, but conduct due diligence on base model and deploy safeguards.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceModerate
DocumentationAdequate
License clarityClear
Deployment complexityHigh
DEV.co fitStrong
Assessment confidenceHigh
Security considerations

No explicit security audit or vulnerability disclosures mentioned. Considerations: (1) Model relies on vLLM/SGLang runtimes; assess those projects' security posture; (2) FP4 quantization reduces data density but does not provide cryptographic guarantees; (3) Tokenizer and tool parsers (glm47, glm45) allow remote code execution if --trust-remote-code is used—verify source before deployment; (4) Base model trained on uncurated internet data may embed prompt-injection or jailbreak vectors; (5) No mention of differential privacy or adversarial robustness; (6) NVIDIA VDP program (https://app.intigriti.com/programs/nvidia/nvidiavdp/detail) available for reporting security issues. Recommendation: Conduct security review of inference framework, implement input validation, and test for prompt injection before production.

Alternatives to consider

zai-org/GLM-5-FP8

Same base model, FP8 precision. ~0.3–1% higher accuracy on benchmarks but ~2× higher memory and slower inference than NVFP4. Choose if accuracy is paramount and GPU memory is abundant.

meta-llama/Llama-3.1-405B or similar (non-quantized)

Larger alternatives with broader community support and fine-tuning examples. Require significantly more VRAM (~500+ GB) and lack MoE efficiency. Consider if NVIDIA-specific ecosystem is not required.

Qwen/Qwen2.5-32B-Instruct or Mistral-Large (quantized)

Smaller, widely-deployed alternatives with more established quantization tooling (llama.cpp, GGUF). Simpler to self-host on limited hardware but lower reasoning/tool-use capability than GLM-5.

Software development agency

Ship GLM-5-NVFP4 with senior software developers

Evaluate GLM-5-NVFP4 for your self-hosted LLM, RAG, or AI agent use case. Start with our vLLM Docker example, verify GPU requirements, and test toxicity/bias guardrails before production. Contact our team to architect multi-GPU tensor parallelism or discuss fine-tuning workflows.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

GLM-5-NVFP4 FAQ

Can I use GLM-5-NVFP4 commercially?
Yes. The MIT license permits commercial use without restriction. The model card explicitly states 'ready for commercial/non-commercial use.' However, ensure the base model (ZAI's GLM-5) has compatible licensing, and implement content safety guardrails—the model inherits internet-crawled biases and may produce toxic outputs.
What GPU do I need to run this model?
NVIDIA Blackwell (B300) or compatible GPUs. Minimum ~100–150 GB VRAM for full model. Typical deployment uses 4–8 GPUs with tensor parallelism (e.g., 8×H100/L40S). CPU inference is not supported. Check vLLM/SGLang documentation for your GPU type's compatibility.
Can I fine-tune GLM-5-NVFP4?
Not documented or supported. The model is post-training quantized; weights cannot be easily updated. For fine-tuning, use the unquantized zai-org/GLM-5 base model and apply LoRA/QLoRA. After fine-tuning, you may re-quantize with ModelOpt if needed.
How much faster is NVFP4 vs. FP8?
Not quantified in the card. NVFP4 uses 4-bit vs. 8-bit precision, so expect ~2× memory savings and faster compute on NVIDIA GPUs. Actual latency gains depend on vLLM/SGLang optimization and batch size. Benchmark your workload with provided vLLM/SGLang examples.

Software developers & web developers for hire

Need help beyond evaluating GLM-5-NVFP4? DEV.co is a software development agency offering software development services and web development for teams of every size. Our software developers and web developers build custom software, web applications, APIs, and open-source llms integrations — and maintain them long-term.

Ready to Deploy GLM-5-NVFP4?

Evaluate GLM-5-NVFP4 for your self-hosted LLM, RAG, or AI agent use case. Start with our vLLM Docker example, verify GPU requirements, and test toxicity/bias guardrails before production. Contact our team to architect multi-GPU tensor parallelism or discuss fine-tuning workflows.