Vision

The Vision API enables NovaAI models to understand and analyze images alongside text. Use it to extract information, answer visual questions, identify objects, describe scenes, and build multimodal AI applications.

Overview

Unlike text-only models, Vision models can process images as input and combine visual understanding with natural language reasoning. Simply provide an image along with an optional prompt, and the model returns a structured response based on what it sees.

Note: Vision models support common image formats including PNG, JPEG, and WebP.

How Vision Works

Rendering diagram…

Analyze an Image

1const response = await client.vision.analyze({
2  model: "nova-vision-1",
3  image: "./workspace.jpg",
4  prompt: "Describe everything visible in this workspace."
5});
6 
7console.log(response.output);

Request Properties

model

string

required

Vision model identifier.

model

file / url

required

Image to analyze.

prompt

string

optional

Additional instructions for the model.

Example Input & Output

Prompt

Image Analyse Prompt

Copy Prompt

Image Analyse Prompt

Copy Prompt

Common Use Cases

Image captioning
Object detection
UI and screenshot analysis
Document understanding
Product image analysis
Visual question answering

Tip: Pair an image with a clear prompt to receive more focused and accurate results.

Next Step

Continue to Files to learn how to upload, manage, and reference files across NovaAI APIs.