Image generation AI 'Stable Diffusion 3' that can output characters correctly is now available via API
The API for the high-quality image generation AI '
Introducing the Stable Diffusion 3 API — Stability AI Japan
https://ja.stability.ai/blog/stable-diffusion-3-api
Stable Diffusion 3 is an image generation AI developed by Stability AI, and human evaluation has confirmed that it is more faithful to prompts than image generation AIs such as DALL-E 3 and Midjourney v6. In addition, a major feature of Stable Diffusion 3 is that it can output characters as instructed by the prompt, allowing you to draw the text you want with the look you want.
High-quality image generation AI 'Stable Diffusion 3' announced, enabling high-precision realization of 'depiction of specified characters' and 'depiction of multiple subjects', which are difficult for image generation AI - GIGAZINE
Stability AI has recently released the APIs for 'Stable Diffusion 3' and 'Stable Diffusion 3 Turbo' on the Stability AI Developer Platform . The APIs are based on credits, with 'Stable Diffusion 3' consuming 6.5 credits and 'Stable Diffusion 3 Turbo' consuming 4 credits per 100 megapixel image. Details of each API can be found at the following links.
Stability AI - Developer Platform
https://platform.stability.ai/docs/api-reference#tag/Generate/paths/~1v2beta~1stable-image~1generate~1sd3/post
Stability AI has also released several examples of Stable Diffusion 3. The image below was generated with the prompt 'A red sofa on top of a white building. Graffiti with the text 'the best view in the city'. (A red sofa on top of a white building. Graffiti with the text 'the best view in the city')', and graffiti is drawn on the wall as instructed by the prompt.
Below is an image generated with the prompt, 'A cardboard box with the phrase 'they say it's not good to think in here', the cardboard box is large and sits on a theater stage. This image also generates characters correctly. It is also characteristic that the cardboard box and background are depicted in detail.
Stability AI will continue to work on improving Stable Diffusion 3 and plan to release model data in the future.
Related Posts:
