Integration

Azure AI Vision integration

Azure AI Vision (formerly Computer Vision) analyzes images given by a public URL with the Image Analysis 3.2 API: sentence captions, tags, object detection with bounding boxes, faces, adult, racy and gory content flags, dominant colors, image type, brands, celebrity and landmark recognition, the most important area of an image for cropping, and printed text extraction (OCR). Uploading raw image bytes, thumbnail generation (binary output), the asynchronous Read operation (its result id comes only in a response header) and Image Analysis 4.0 are not exposed. Authentication is a key of your Azure AI Vision or multi-service resource sent in the Ocp-Apim-Subscription-Key header; the base URL uses the resource's custom subdomain (https://<resource-name>.cognitiveservices.azure.com).

Connect Azure AI Vision How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Azure AI Vision.Built from the 8 Azure AI Vision tools below, on the connection you add once.

What you can do with Azure AI Vision

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Image, 4 operations

analyze_imageAnalyze imagePOSTAnalyzes an image URL (Computer Vision 3.2) and returns the features listed in visualFeatures: categories, tags, description, faces, image type, color, adult content, objects, brands.
analyze_image_by_domainAnalyze image by domainPOSTRecognizes celebrities or landmarks in an image URL with a domain-specific model; pass a model name from list_models (celebrities or landmarks).
describe_imageDescribe imagePOSTGenerates human-readable sentence captions of an image URL, ordered by confidence, plus the content tags they are based on. maxCandidates sets how many captions come back (default 1).
get_area_of_interestGet area of interestPOSTReturns a bounding box around the most important area of an image URL, useful to crop it.

Model, 1 operation

list_modelsList modelsGETLists the domain-specific models usable with analyze_image_by_domain (today celebrities and landmarks) with their categories.

Object, 1 operation

detect_objectsDetect objectsPOSTDetects objects in an image URL and returns each object's label, confidence, bounding rectangle and parent object in the taxonomy.

Tag, 1 operation

tag_imageTag imagePOSTReturns content tags (objects, living beings, scenery, actions) for an image URL with confidence scores; some tags carry a hint giving context.

Text, 1 operation

recognize_printed_textRecognize printed textPOSTSynchronous OCR of printed text in an image URL, returned as regions, lines and words with bounding boxes (bounding boxes are comma-separated left,top,width,height strings).

Use Azure AI Vision from Claude Code, Cursor or Codex

Connect Azure AI Vision in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Azure AI Vision operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Azure AI Vision

Create your workspace
Sign up free. Nothing to install.
Connect Azure AI Vision once
Add your Azure AI Vision connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.