AI
New top local AI image editor is here!AI Search12 Nov 2026
Mentioned in the conclusion alongside Krea 2 as better than MiniMax H3 in some aspects of text-to-image.
Open-weights 5B text-to-image model family from OpenMOSS with native JSON bounding-box layout control and text rendering.
Updated EU status checked Last shown
MOSS-Image at a glance
- What is MOSS-Image?
- MOSS-Image is a model from OpenMOSS. Open-weights 5B text-to-image model family from OpenMOSS with native JSON bounding-box layout control and text rendering.
- Who makes MOSS-Image?
- MOSS-Image is made by OpenMOSS.
- Does MOSS-Image work in the EU?
- Available in the EU. Weights are downloadable from Hugging Face under Apache-2.0 with no regional restriction, so they can be run locally in the EU/EEA. Last verified 14 Nov 2026.
- How much does MOSS-Image cost?
- Open source. Weights free to download under Apache-2.0; run locally with your own GPU. Licence: Apache-2.0 (weights).
- Where was MOSS-Image shown?
- In 3 creator videos by AI Search. Most recently by AI Search on 12 Nov 2026, at 19:25 in “New top local AI image editor is here!”.
Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
About
MOSS-Image is a text-to-image model family from the OpenMOSS Team, released as a base checkpoint (MOSS-Image-Base) and a fine-tuned checkpoint (MOSS-Image-SFT) trained on curated data. Both share the same 5B-parameter architecture and sampling interface. Prompts can be up to 4K tokens, and the model is described as handling text rendering as well as anime and photorealistic styles.
Beyond natural-language prompts, it supports native layout control. A structured JSON prompt specifies the scene, objects, rendered text and bounding boxes for each element. The project also provides a Prompt Enhancer skill for agents that turns ideas into captions or layout JSON. Community add-ons listed on the ComfyUI repackage include a color/depth control LoRA and an image-edit LoRA. The base model itself is text-to-image only.
It is accessed through a Diffusers pipeline (Python 3.10+, NVIDIA GPU), a Gradio app with several memory modes for lower-VRAM setups, or ComfyUI, using a native plugin and repackaged files from Comfy-Org (bf16 and int8 variants). Model weights are Apache-2.0; the pipeline code and third-party components keep their own licenses. It was announced on 2026-10-16.
Key features
- 5B-parameter text-to-image backbone with prompts up to 4K tokens
- Native layout control via JSON prompts with bounding boxes
- Text rendering, including Chinese text, plus anime and photorealistic styles
- Base and SFT checkpoints sharing one interface
- Diffusers, Gradio and ComfyUI support with memory-saving modes
- Apache-2.0 weights; supports LoRA fine-tuning
3 moments
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
New top local AI image generator is here!
Open-source 5B-parameter text-to-image model with strong knowledge, text rendering, infographics, and layout control via bounding boxes. Runs on low VRAM; creator installs it in ComfyUI, tests text-to-image and layout workflows, and notes it is text-to-image only (inpainting for editing).
AI
Quantum breakthrough, Gemini 4 Argon, MOSS Image, KLING 4, OpenAI spy, Muse Wonder: AI NEWSAI Search18 Oct 2026
Open 5B text-to-image model under Apache 2, relatively unfiltered, good at anime, game scenes and text rendering, with bounding-box layout control. Released as a base model and an SFT model; no editing.
Keep exploring
Similar AI tools
Closest neighbours by meaning, not just by category.
ZI
Z-Image
Model
Open-weights 6B text-to-image model family from Alibaba's Tongyi-MAI lab, with a fast 8-step Turbo variant.
Available in EUImage Generation
26 Nov 2025Open source
KR
Kromoss
Model
Community full fine-tune of MOSS-Image-Base at 1024 px, with optional 8-step turbo LoRAs; open weights under Apache-2.0.
Available in EUImage Generation
19 Oct 2026Open source
BI
Boogu Image
Model
Open image generation and editing model family (base, turbo, edit) released under Apache 2.0.
Available in EUImage Generation
Release date n/aOpen source
F1
Fibo 1.5
Model · Bria AI
Open-weights text-to-image model from Bria AI, fine-tuned from FIBO, supporting structured JSON prompts and image editing.
Available in EUImage Generation
13 Jul 2026Open source
MI
MOSS Image 8-step DMD LoRA
Model
8-step distillation LoRA for MOSS-Image-SFT that runs without CFG, for ComfyUI.
Available in EUImage Generation
16 Oct 2026Open source
LL
LLaDA-Image
Model · inclusionAI
Open-weights 6B unified text-to-image and instruction-based image editing model family from inclusionAI.
Available in EUImage Generation
4 Sept 2026Open source