zen5-nano-9B-gguf

zenlm/zen5-nano-9B-gguf

zen5-nano-9B-gguf is a image-text-to-text model from zenlm. Use it for the image-text-to-text task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as apache-2.0.

Edge / on-device tier of the Zen5 family. Multimodal dense vision-language model (image + text in - text out) at the 9B scale, qwen35 architecture, tuned for fast inference on laptops, phones, and consumer GPUs.

Downloads · 30 days

6% of all-time downloads

All-time downloads

281

Public

Parameters

9.7B

19.3 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors19.3 GB · 100%

Parameter types

How the weights are stored.

BF169.7B · 100%

Try a prompt

Base models

Qwen/Qwen3.5-9B

Model card

From the Hugging Face model README

Zen5 Nano 9B

Edge / on-device tier of the Zen5 family. Multimodal dense vision-language model (image + text in -> text out) at the 9B scale, qwen3_5 architecture, tuned for fast inference on laptops, phones, and consumer GPUs.

Repackaged from Qwen/Qwen3.5-9B (apache-2.0, Alibaba Qwen). Native HuggingFace safetensors (Qwen3_5ForConditionalGeneration), re-hosted under the Zen5 nano line. Not trained from scratch — a permissively-licensed redistribution for the OSS-clean Zen model line.

Zen5 nano ladder — pick by hardware budget:

SKU	Params	Base	Fit
`zen5-nano-0.8B`	0.8B	Qwen3.5-0.8B	Raspberry Pi 5 / phone CPU / browser WASM
`zen5-nano-2B`	2B	Qwen3.5-2B	low-end iGPU / 8 GB RAM laptop
`zen5-nano-4B`	4B	Qwen3.5-4B	16 GB RAM laptop / mobile NPU
`zen5-nano-9B`	9B	Qwen3.5-9B	24 GB+ unified RAM / consumer GPU

The larger Zen5 tiers (zen5-flash, zen5, zen5-pro, zen5-max, zen5-coder) cover everything from agentic-default to frontier-quality multi-GPU.

Files

Native HuggingFace safetensors. Load with transformers following the upstream Qwen/Qwen3.5-9B recipe, or fetch directly:

hf download zenlm/zen5-nano-9B-gguf --local-dir zen5-nano-9B-gguf

Run

Hosted via the Hanzo gateway (api.hanzo.ai) as zen5-nano-9B.

License

apache-2.0. Upstream: Qwen/Qwen3.5-9B by Alibaba Qwen. This repository redistributes the weights under the same license.

Task: Image-Text-to-Text
Library: transformers
Type: qwen3_5
License: apache-2.0
Languages: en, zh
Created: May 30, 2026
Updated: Jul 14, 2026