API
Vision API
Raw HTTP endpoints for AI image analysis, tagging, embeddings, and editing
Vision API
Raw HTTP endpoints for CoreViz's AI vision capabilities. If you're using JavaScript or TypeScript, the @coreviz/sdk npm package is the recommended way to access these endpoints — it handles authentication, retries, and TypeScript types.
Base URL
https://lab.coreviz.io/api/aiAuthentication
All Vision API endpoints require an API key:
-H "x-api-key: YOUR_API_KEY"Get your key from Your Account → API Keys on lab.coreviz.io.
Endpoints
| Method | Endpoint | SDK Method | Description |
|---|---|---|---|
POST | /api/ai/describe | sdk.describe() | Generate image description |
POST | /api/ai/tag | sdk.tag() | Tag / classify image |
POST | /api/ai/embeddings | sdk.embed() | Generate CLIP embedding |
POST | /api/ai/edit | sdk.edit() | Edit image with prompt |
POST | /api/ai/generate | sdk.generate() | Generate image from text |
Supported Image Formats
- JPEG, PNG, WebP, GIF
- Images must be publicly accessible via HTTPS
- Maximum file size applies (check your plan)
Error Handling
| Status | Description |
|---|---|
400 | Invalid request (missing fields, invalid image URL) |
401 | Invalid or missing API key |
422 | Image could not be processed |
429 | Rate limit exceeded |
500 | Server error |