File	Role	Notes
`Qwen3-VL-4B-Instruct-Unredacted-MAX.BF16.gguf`	Language model (BF16)	~8 GB
`Qwen3-VL-4B-Instruct-Unredacted-MAX.mmproj-bf16.gguf`	Vision projector (mmproj)	Required for image understanding

ComfyUI (H3 Prompt Writer)

Place both files under:

ComfyUI/models/llm_gguf/

In H3 Prompt LLM Loader (GGUF):

model_name → the .BF16.gguf
mmproj_name → the .mmproj-bf16.gguf

Attribution / lineage

Base model: Qwen/Qwen3-VL-4B-Instruct (Apache-2.0)
Unredacted / abliterated finetune: prithivMLmods/Qwen3-VL-4B-Instruct-Unredacted-MAX
GGUF source pack: prithivMLmods/Qwen3-VL-4B-Instruct-Unredacted-MAX-GGUF

This repo redistributes those GGUF weights under Apache-2.0 for convenience. All credit to the original authors; this is a rehost with a shorter community name.

License

Apache License 2.0 — same as the upstream Qwen3-VL and Unredacted-MAX releases.

EllipsesMark/qwen3-vl-4b-unleashed

qwen3-vl-4b-unleashed

Files

ComfyUI (H3 Prompt Writer)

Attribution / lineage

License

EllipsesMark/
qwen3-vl-4b-unleashed