• PREDICTIONS
  • VALIDATION
  • DATA
  • HUMAN

How to run model inference with your AI client

Connect Roboflow to your AI client once. From then on, you can test model accuracy by sending images for real-time inference without leaving your workspace.

Ask AI about this page

Short answer

How can I run model inference with my AI?

You can trigger inference on specific images by telling your AI to run a model from a Roboflow project. This lets you verify how your model handles new data immediately. You'll get the prediction results right in your chat or IDE.

The outcomes

What you get from inference

Running inference turns your AI into a testing tool for your computer vision models.

PREDICTIONS

Real-time bounding boxes

The AI runs the model on your image and returns the detected objects and coordinates. You see exactly what the model sees.

VALIDATION

Instant accuracy checks

Compare model outputs against your expectations immediately. This helps you spot false positives or misses without switching tabs.

DATA

Project context

The AI pulls specific version details to ensure it uses the right model weights for the task.

HUMAN

Manual review needed

The AI provides the raw inference data, but you decide if the model's performance meets your requirements for deployment.

The workflow

What your AI does when the image arrives.

The AI handles the heavy lifting of communicating with Roboflow to get your results.

  1. Locate the project

    The AI identifies the correct Roboflow project you want to work with.

    get_project

  2. Select the version

    It finds the specific model version containing the weights you need for this test.

    get_version

  3. Execute inference

    The AI sends the image data to the model to generate predictions.

    run_inference

  4. Report results

    The AI presents the model's findings, such as object classes and confidence scores, in your chat.

    run_inference

Try it

Copy these to start.

Use these prompts to kick off your first inference task.

Starting points

You'll need to replace the project names and version IDs with your actual Roboflow details.

Roboflow Connector

You're all set. Choose your MCP client and follow the setup instructions.

Connector linkhttps://edge.vinkius.com/vk_preview_3PcAB3lxgco6m9RlryrdB51L1KU7rDlnqGHOwOaP/mcp

Claude Desktop

Follow the steps below to connect in seconds.

  1. 1In Claude Desktop, open Settings → Connectors.
  2. 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
  3. 3Click Add and start a new chat — Roboflow capabilities are ready to use.
Configuration · claude_desktop_config.jsonCopy
{
  "mcpServers": {
    "roboflow-mcp": {
      "url": "https://edge.vinkius.com/vk_preview_3PcAB3lxgco6m9RlryrdB51L1KU7rDlnqGHOwOaP/mcp"
    }
  }
}
Copy into chat04

  • Run inference on this image using the 'traffic-detection' project version 2.

  • Check the 'warehouse-safety' project version 5 with this image and tell me what you find.

  • Use the model from project 'crop-health' version 1 to analyze this photo.

  • Run inference on this image using the latest version of my 'defect-detection' project.

  • Claude
  • ChatGPT
  • Cursor
  • VS Code
  • Windsurf
  • Claude Code
  • JetBrains
  • Cline

Start here

Connect Roboflow once, then ask.

Just link your account through the Vinkius dashboard. Your credentials stay encrypted, and you can start running inference in your preferred IDE or chat client immediately.

Connect Roboflow to your AI

FAQ

How this task behaves

  • 01

    Can the AI upload images to Roboflow for me?

    No. This connector is for running inference on existing images, not for uploading new data to your datasets.

  • 02

    Does the AI change my model settings?

    No. The AI can only trigger inference using existing versions; it cannot modify your model training or project configurations.

  • 03

    Can I run inference on a specific version of a model?

    Yes. You can specify which project and version you want to use for the inference task.

  • 04

    Will the AI automatically save the results to my project?

    No. The AI brings the inference results back to your chat or IDE for you to review.

  • 05

    Do I need to provide the image every time?

    Yes. You need to provide the image or the image path so the AI has something to run the model against.

  • More questions about Roboflow? See everything the Roboflow Connector can do

Connect Roboflow to Claude, Cursor, ChatGPT & more