Open-Source
Use cases
Gallery
Grow with us
Model library
Use any open model, or import your own as a baseline for your fine-tune.
Check the model galleryAbstract
Single API call for fine-tuning, batch of predefined prompts, upscaling, face correction and much more.
GuidesNo devops. Handle peaks
Unique infrastructure adjusted for fine-tuning with auto-scaling. No queues. Fast processing.
API overviewFAQ
Fine-tuning is the process of adapting a pre-trained generative image models, like Flux, Stable Diffusion 1.5 (SD1.5) or Stable Diffusion XL (SDXL), to create personalized results for specific subjects or styles. Astria's fine-tuning API allows generating studio-quality images of a subject, including AI-generated professional headshots, as well as cartoons and avatars.
Astria offers a FaceID like / adapter alternative approach alongside the Dreambooth fine-tuning option, which results in lower resemblance but requires significantly less processing time. You can read more about this option in the docs. This is more suitable for creating lower fidelity results. We provide different types of fine-tuning, such as Checkpoint and LoRAs, utilizing both fine-tuning for SDXL and SD1.5 technologies.
Flux is a family of generative image models created by Black Forest Labs allowing a new level of text to image generation. See more in the documentation.
Yes! Customers may use the playground GUI and API services services without any additional licensing requirements.
Generative filters use generative image models in order to achieve a new level of control and effects to images. Use Astria's inference API to create a rich variety of artistic filters such as line-art, oil painting effects or unique artistic illustration together with controlnet and fine-tuning while preserving your custom subject as as person identity. Astria's offers advanced options which can help you achieve artistic styles, perfect for enhancing both realistic and more artistic images.