Model Detail

GLM-5.3-Flash-NVFP4 is an AI model with 84.6B parameters released by local-inference-lab. It has accumulated 3K downloads on Hugging Face since publication.

GLM-5.3-Flash-NVFP4 ships with 84.6B parameters. Total weight footprint is approximately 169.1 GB, which is the relevant figure when planning local-inference VRAM.

GLM-5.3-Flash-NVFP4 is best fit for general-purpose AI workloads. Treat this as a starting matrix rather than a benchmark verdict — the right deployment usually depends on the specific evaluation suite that mirrors your workload.