Midjourney is bought for visuals: ads, social posts, covers. Open image models do the same work on your own machine or server: you can generate as much as you like, prompts and reference images never leave your side, and a model can be adapted to your style or product with a separate fine-tune. In exchange you take on the GPU, the interface setup and the search for good settings — a usable picture rarely comes out of the first prompt, it usually takes several iterations. Composition, hands and text inside the image are still uneven in open models, so part of the output has to be discarded or retouched. The one thing to check before you start is the license of the exact version: within a single family it can be either permissive or strictly non-commercial.
Sber's Russian family of image and video generation models. Understands Russian-language prompts and Russian cultural context well; released under MIT.
Image generation from the creators of Stable Diffusion. Renders text in images well and keeps the composition.
The model that started open image generation. A huge ecosystem of fine-tunes, styles and plugins; runs even on a home PC. The popular SDXL-Lightning and Hyper-SD accelerators were made by ByteDance.
Image generation and editing, including text in images. Earlier versions allow commercial use; the latest 2.1 is non-commercial only.
A compact 6B model with photorealism on par with large models. The Turbo version produces an image in a few steps on a regular gaming GPU.
Kandinsky and Z-Image are Apache 2.0 and MIT, whereas FLUX [dev] versions are non-commercial, Stable Diffusion licensing depends on the generation, and the recent Qwen-Image-2.1 was released under a research license. For commercial campaigns, have the exact version and its license reviewed by a lawyer.
An open model can run on your own server: data stays in-house, there is no per-request fee, and the model can be fine-tuned on your documents.
