Nara Lens
Now in Public Beta

ProgrammablePerception

Computer vision APIs that see what you tell them to see. Define perceptions in natural language, get structured data back.

Get StartedView API Reference
Bash
curl -X POST https://api.naralens.com/v1/analyze \
  -H "Authorization: Bearer $NARA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image": "https://example.com/image.jpg",
    "perception": "detect:person,vehicle"
  }'

Get Started in Minutes

Choose your preferred SDK or use the REST API directly. All methods provide the same powerful perception capabilities.

Python SDK

pip install naralens

from naralens import Vision

client = Vision(api_key="...")
result = client.analyze(
    image="image.jpg",
    perception="count:cars"
)
Get Started

JavaScript SDK

npm install naralens

import { NaraLens } from 'naralens';

const nara = new NaraLens(apiKey);
const result = await nara.analyze({
  image: imageUrl,
  perception: 'detect:face'
});
Get Started

REST API

cURL examples & OpenAPI

curl -X POST /v1/analyze \
  -H "Authorization: Bearer $KEY" \
  -d '{"image": "...",
       "perception": "..."}'
Get Started
Capabilities

Everything You Need for Computer Vision

From simple object detection to complex scene understanding, Nara Lens provides the building blocks for any vision application.

Object Detection

Core

Detect and classify thousands of object types with state-of-the-art accuracy. Returns bounding boxes, confidence scores, and labels.

Semantic Segmentation

Core

Pixel-level understanding of images. Separate foreground from background, identify regions, and create precise masks.

Natural Language Queries

New

Define what you want to see using plain English. Ask "count the red cars" or "find all faces looking left".

Video Streaming

Beta

Real-time analysis of video streams via WebSocket. Process frames with sub-100ms latency for live applications.

Custom Perceptions

Enterprise

Train custom perception models on your data. Create specialized detectors for unique use cases without ML expertise.

Batch Processing

Core

Process thousands of images in parallel with our batch API. Optimized for large-scale data processing workloads.