Model Detail

GLM-5.3-NVFP4 is a large language model with 195.5B parameters released by incoai. The model is registered under the text-generation pipeline tag on Hugging Face, distributed under a other license.

GLM-5.3-NVFP4 ships with 195.5B parameters. Total weight footprint is approximately 390.9 GB, which is the relevant figure when planning local-inference VRAM. Distribution is governed by the other license — review the exact terms before commercial deployment.

GLM-5.3-NVFP4 is best fit for general-purpose chat and instruction-following 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.