--- inference: true language: - en tags: - stable-diffusion - text-to-image license: creativeml-openrail-m --- # Seek.art MEGA is a general use model that significanlty improves up on 1.5 across dozens of styles, by [seek.art](https://seek.art/) This model was trained on nearly 10k high-quality public domain digital artworks with the goal of improving output quality across the board. We find the model to be highly flexible in its ability to mix various styles, subjects, and details. We recommend resolutions above 640px in one or other dimensions for best results. You can try this model and several others for free at [seek.art](https://seek.art/). We also recommend an inference tool supporting prompt weighting and high resolution optimization / fixing for best results. We suggest [InvokeAI](https://github.com/invoke-ai/InvokeAI) as a sensibly licensed and fully featured open-source inference tool. ### Examples All the above example images including all relevent settings are available [here](https://seek.art/explore/search?collection=6112a64d-bd8b-4043-8d96-88c7cfa65c43). Additionally, search thousands of high quality prompts on seek.art for free.