Model Detail

supergemma4-26b-abliterated-multimodal-mlx-4bit is a multimodal model with 26B parameters released by Jiunsong. The model is registered under the image-text-to-text pipeline tag on Hugging Face, released under the gemma license.

supergemma4-26b-abliterated-multimodal-mlx-4bit ships with 26B parameters. Total weight footprint is approximately 4.6 GB, which is the relevant figure when planning local-inference VRAM. Distribution is governed by the gemma license — review the exact terms before commercial deployment.

supergemma4-26b-abliterated-multimodal-mlx-4bit is best fit for mixed text-and-image reasoning tasks such as document understanding. 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.