Source: factory.ai/ObjAiVision.py
Producer/consumer pipeline orchestrator for image analysis.
Connects any image source (URL capture, file path, or raw base64) to
ObjAI's vision model consumer interface. Deliberately agnostic about
which model handles each end of the pipeline.
...
| Method | Signature | Description |
|---|---|---|
| analyse_url | analyse_url(url: str, prompt: str = '', role: str = '') -> str |
Capture url via the configured provider, then analyse with the |
| analyse_image | analyse_image(image_path: str, prompt: str = '', role: str = '') -> str |
Analyse an existing image from a file path or base64 string. |
| analyse_url_diff | analyse_url_diff(url: str, prompt: str = '', role: str = '') -> str |
Capture url, compare with the most recent stored analysis for that |
| analyse_image_multi | analyse_image_multi(image_path: str, models: list[str] | None = None, prompt: str = '', role: str = '') -> dict[str, str] |
Analyse an image with multiple vision models and return all responses. |