API Updates

Image Prompt limit increased

The prompt limit on image generation has been increased to 1,500 characters on supported models (e.g. Flux and SDXL based models).

Image Generation: Safe Mode

Image generation now supports the filtering of adult content by passing in the safe_mode paramter as true. Safe mode will screen for adult content, and if detected, will blur the image and set the x-venice-is-blurred header to true. The docs have been updated and an example Postman request can be found here.

Image Generation: Stable Diffusion 3.5 Support

Stable Diffusion is now supported in the image generation endpoint.

Image Styles API

An API endpoint has been updated to list currently supported API styles. PLease see the docs here.

Function Calling Support

Function calling support has been introduced using the Llama 3.3 70B model. Use the function_calling_default model trait. See an example request within Postman here.

Pony Realism activate in image generation API

The pony-realism model has been activated within the image generation API. It is now tagged with the most_uncensored trait.

Model Filtering

The /api/v1/models endpoint has been updated to support filtering by a type parameter. Check out the update here.

New Docs ✨

We’ve redesigned and updated our docs to include a new API reference section. This section includes all of the available endpoints and their parameters.