Skip to content

API Integration

This guide shows how to integrate the Raku Sight Inference API into your application. The API provides image inference using YOLO and LLM-based models, with support for both batch and real-time streaming responses.

Base URL

https://sightapi.raku.so/api/v1

Replace this with your self-hosted instance URL if applicable.

Authentication

All endpoints require an api_key. Depending on the endpoint version:

  • v1 endpoints — pass api_key as a multipart/form-data field alongside the file upload.
  • v2 endpoints — pass api_key inside the JSON request body.

Keep your API key secret

Never expose your api_key in client-side code or public repositories. Use environment variables or a secrets manager.

Endpoints

Method Path Description
POST /inference/predict/batch Batch prediction with file uploads (v1)
POST /inference/predict/stream Streaming prediction with file uploads (v1)
POST /inference/predict/batch/v2 Batch prediction with image URLs (v2)
POST /inference/predict/stream/v2 Streaming prediction with image URLs (v2)
POST /inference/assets/signed-url Generate a signed URL for a result asset

Response Codes

Code Meaning
200 Success
400 Bad request — check your parameters
500 Server or upstream service error

Data Types

DataSchema (v2 endpoints)

Field Type Description
image_url string Public URL of the image to process
filename string Filename for identification
is_yolo boolean Run YOLO object detection
is_llm boolean Run LLM-based analysis

PredictSchema (v2 request body)

Field Type Required Description
api_key string Yes Your API key
query string Yes Natural language query for LLM analysis
data DataSchema[] Yes List of images to process