Back to Model
Q

Qwen3.5-4B Uncensored Browser GGUF

Model Nov 3, 2026

InterpreticAI's 1.86 GB IQ3_XXS quantization targets browser and mobile inference, with author-reported losses in general capability and more degenerate answers.

2 sources attached
self-hostedmultimodal

What happened lately

Qwen3.5-4B Uncensored Browser GGUF timeline

InterpreticAI released this browser-oriented IQ3_XXS build on November 3, 2026. The text GGUF is 1,864,652,736 bytes; the optional image projector is 672,423,424 bytes. These are download sizes, not demonstrated runtime memory requirements.

The English card targets in-browser and on-device inference and provides a llama.cpp command. Its own tests report reduced general capability and more degenerate answers than the larger builds. The parent modification targets political censorship while retaining most harmful-request refusals in the author’s test. AITrending verified the release files and download availability, but did not run inference or reproduce the evaluations.