Changelog
Follow up on the latest improvements and updates.
RSS
fixed
improved
Separate server for FaceGen
Facegen is a bit slow and takes 25 seconds to generate an image. This is due to the model getting loaded repeatedly on the memory. that consume lots of time. Having a single server just for faceGen Will reduce the Image generation time.
new
improved
Updates 7.12
- FaceGen now has a style parameter.Now you can add Anime style, 3D style, or comic style and get out put according to the style that has been added.
- Added a new pipeline for getting image generation in secondsand getting better image quality compared to before.
- We have replaced Inf Edit with object remover:Users can create a mask image around the object and remove that object
- New and improved version of Uncensored Chat:Can do better RolePlay, stick to the system prompt.
- Instant ID:Just like headshot but now the generated image can have a different style. We have added support for different styles for image generation.
A curation of high-quality generated images, videos, GIFs, audio's and more. Interact with the community feed, upvote, get to know the prompts and more.
Join our affiliate program and start earning commissions for your referrals.
Help your network learn more, build more on AI, and get paid for it. Learn more by signing up and checking out your dashboard - https://modelslab.com/
- Added a new pipeline for getting image generation in seconds.
- Users can create a mask image around the object and remove that object
Users Can access all the APIs available on the Platform with 39,999 API calls.
Our latest update includes
30 API calls
before you are prompted or tasked with subscribing. Test out the capabilities (we are constantly improving and customising, so results may vary).We wanted to reduce the barrier to the entry. Start playing with our playground today or through API calls.
All generated images will be auto deleted in 30 days.
Additionally, you can deploy some S3 buckets on your end to store your old generated images.
All your generations will be 100% private and not be used in training any model.
SD3 is now available for everyone.
SD3 excels at a variety of styles, and interprets prompts with far greater accuracy than previous base models. It's also an incredible step forward in text adherence.
Access: "model_id": "stable-diffusion-3-medium"