Skip to content

AI Backends

Visionati aggregates multiple AI services. Each backend has different strengths and costs.

These generate natural language descriptions of your images.

BackendCompanyCost
claudeAnthropic3 cents/description
openaiOpenAI2 cents/description
grokxAI2 cents/description
geminiGoogle1 cent/description
jinaaiJina AI1 cent/description
llavaReplicate1 cent/description
bakllavaReplicate1 cent/description

These detect tags, faces, colors, text, brands, and NSFW content.

BackendTagsNSFWFacesColorsBrandsTexts
clarifai0.15 cents0.15 cents
googlevision0.2 cents0.2 cents0.2 cents0.2 cents0.2 cents0.2 cents
imagga0.15 cents0.15 cents
rekognition0.15 cents0.15 cents0.15 cents

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.

bakllava, clarifai, claude, gemini, googlevision, grok, imagga, jinaai, llava, openai, rekognition