AI Backends
Visionati aggregates multiple AI services. Each backend has different strengths and costs.
Description Backends (LLMs)
Section titled “Description Backends (LLMs)”These generate natural language descriptions of your images.
| Backend | Company | Cost |
|---|---|---|
claude | Anthropic | 3 cents/description |
openai | OpenAI | 2 cents/description |
grok | xAI | 2 cents/description |
gemini | 1 cent/description | |
jinaai | Jina AI | 1 cent/description |
llava | Replicate | 1 cent/description |
bakllava | Replicate | 1 cent/description |
Detection Backends (Computer Vision)
Section titled “Detection Backends (Computer Vision)”These detect tags, faces, colors, text, brands, and NSFW content.
| Backend | Tags | NSFW | Faces | Colors | Brands | Texts |
|---|---|---|---|---|---|---|
clarifai | 0.15 cents | 0.15 cents | — | — | — | — |
googlevision | 0.2 cents | 0.2 cents | 0.2 cents | 0.2 cents | 0.2 cents | 0.2 cents |
imagga | 0.15 cents | 0.15 cents | — | — | — | — |
rekognition | 0.15 cents | 0.15 cents | 0.15 cents | — | — | — |
Defaults
Section titled “Defaults”When no backend parameter is specified, the following backends are enabled:
Claude, Gemini, Grok, OpenAI, Clarifai, Google Vision, Rekognition
To use a different set, pass the backend or backend[] parameter with the values you want. Jina AI, LLaVA, BakLLaVA, and Imagga are available but not enabled by default and must be explicitly requested.
All Valid Backend Values
Section titled “All Valid Backend Values”bakllava, clarifai, claude, gemini, googlevision, grok, imagga, jinaai, llava, openai, rekognition