Model Detail

MLX-Qwopus3.5-27B-v3-4bit is a large language model with 27B parameters released by Jackrong. The model is registered under the text-generation pipeline tag on Hugging Face, distributed under the permissive apache-2.0 license.

MLX-Qwopus3.5-27B-v3-4bit ships with 27B parameters. Total weight footprint is approximately 26.9 GB, which is the relevant figure when planning local-inference VRAM. The apache-2.0 license is permissive, allowing commercial deployment and derivative work without per-seat fees, though attribution requirements still apply.

MLX-Qwopus3.5-27B-v3-4bit 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.