agentai-image

Image generation and editing for AgentAI.

• ImageGenerator generates and edits images from your code.
• With the toolbox feature (on by default), GenerateImageToolBox gives an agent the generate_image and edit_image tools, with a required file name that is passed to the storage. The model gets the image, or only its metadata as JSON. ReadMediaToolBox lets the model read a stored file by its ID.
• Images are saved in any MediaStorage (MediaRead + MediaSave, MemoryStorage) from agentai_image::storage, re-exported from agentai-core.
• TLS features as in agentai: rustls-ring (default), rustls-aws-lc, native-tls, system-proxy.

use agentai_image::ImageGenerator;

// The provider comes from the model name, the API key from `OPENAI_API_KEY`
let generator = ImageGenerator::new("openai::gpt-image-1")?.with_size("1024x1024");
let response = generator.generate("A lighthouse at dawn").await?;

Providers live in agentai_image::provider. provider::openai speaks the OpenAI image API and any API compatible with it (OPENAI_BASE_URL, or a base URL set in code); provider::openrouter speaks the OpenRouter image API (openrouter::<model>, key from OPENROUTER_API_KEY). provider::list_models("openai" | "openrouter") lists the image models a server offers. See the image_generator, image_tools and list_models examples.