Skip to content
MCP server

Dialect demo

Turns an existing RTSP or ONVIF camera into a live API that agents and apps can query

Answered our latest tool-list check 3 hours ago with 4 tools. It will be marked verified after 3 successful checks in 7 days.

Install
Pricing
Paid
Platforms
API, Web, self-hosted
Runs
Remotely
Version
1.0.0
Listing updated
1 day ago

Similar MCP servers

About Dialect demo

Dialect demo is a paid MCP server for physical world apis. It runs as a hosted remote server, so there is nothing to install locally.

The Dialect site describes a camera-to-API layer: you state in one sentence what a camera should watch, confirm the proposed reading, and then read typed values from it. Each question becomes its own endpoint on the same camera, and every value comes with a confidence level and the time since it last changed.

Dialect runs in Docker on a computer placed beside the camera. The site says the setup can be started by pasting a prompt into an agent, which locates the camera and starts the service. Values are delivered through an MCP tool, a REST call, or a signed webhook that fires on each change, so agents can ask questions the original setup did not anticipate.

The product is aimed at teams that want software or agents to act on what a physical space looks like, such as aviation hangars, test chambers or utility equipment. A public, read-only demo endpoint is available without an account, and the paid plan is priced per question on the camera.

Install

Run in a terminal

claude mcp add --transport http dialect-demo https://api.dialectlayer.com/mcp/public

Add to Settings → Connectors → Add custom connector

https://api.dialectlayer.com/mcp/public
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "dialect-demo": {
      "url": "https://api.dialectlayer.com/mcp/public"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "dialect-demo": {
      "type": "http",
      "url": "https://api.dialectlayer.com/mcp/public"
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "dialect-demo": {
      "serverUrl": "https://api.dialectlayer.com/mcp/public"
    }
  }
}

Add to Grok Bot desktop app → Connect apps → Add a custom connector

https://api.dialectlayer.com/mcp/public

Enter a Name (for example "Dialect demo"), choose Type: Remote HTTPS, paste this as the Server URL and choose Add plugin.

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "dialect-demo": {
      "type": "http",
      "url": "https://api.dialectlayer.com/mcp/public"
    }
  }
}

Using several servers? Build one config file with the MCP config generator.

Tools this server provides (4)

Read directly from the server.

All 4 tools say they're read-only.

From the server's own tool annotations; not verified by us.

dialect_endpointsSays read-only
Use first: the endpoints anyone may read, each with its name and title. Side effects: none.
dialect_read_endpointSays read-only
Use to read a public endpoint now: each field's question, value, whether it was read sure, why not, and since when (Unix seconds) it has held, plus the last 24 hours summed and its latest changes. A value read unsure is what it leaned to. Side effects: none.
dialect_summarySays read-only
Use for how long or how often over a time: each value's sure seconds, share of the window and how many times it was entered, over the last last_s seconds (default a day). Side effects: none.
dialect_changesSays read-only
Use to see what changed: each reading that changed a field, newest first, with its value, whether it was read sure and when, over the last last_s seconds (default an hour). Side effects: none.

What it does

Sentence-based endpoint setup
Each endpoint is defined by describing the observation in a sentence and confirming the proposed setup, with no model training.
Typed values with confidence
Readings include a value, how sure Dialect is about it, and how long it has held.
MCP, REST and webhooks
Results can be read through an MCP tool, a REST endpoint, or pushed as signed webhook events when values change.
Works with RTSP and ONVIF cameras
The site says any camera speaking RTSP or ONVIF can be connected, running through a Docker service on a nearby computer.
Reports unclear readings
When the view is blocked, the reading shows a covered state instead of guessing, and it recovers when the view clears.

Good for

  • Letting an AI agent check whether a hangar bay or equipment area is occupied
  • Triggering an alert in an app when a monitored valve or indicator changes state
  • Keeping a daily history of a physical state for later questions from an agent
  • Monitoring a camera mounted on moving equipment without reconfiguring each time it shifts

Pricing

Paid, from $99 a month

$99/mo for one question, $49 per additional question; first answer free, 30-day refund

Prices change. Check dialectlayer.com before you buy.

Dialect demo FAQ

What is Dialect?

Dialect turns a camera into a live API. You describe what it should observe in a sentence and confirm what it proposes. Your software and agents then read typed values, with how sure each one is and since when it has held. Nothing is trained for any endpoint.

Which cameras does Dialect work with?

Any camera that speaks RTSP or ONVIF. Dialect runs in Docker on a computer beside it.

Do I have to train a model?

No. Every endpoint is set up from a sentence, and nothing is trained for it.

How much does Dialect cost?

$99 a month, with $49 for each question after the first. Your first answer is free and there is a 30-day full refund. Custom pricing applies to more than 10 questions or to security and contract needs.

Is Dialect demo free?

No. Dialect demo is a paid product. Paid plans start at $99 a month.

How do I add Dialect demo to Claude Code?

Run this command in your terminal: claude mcp add --transport http dialect-demo https://api.dialectlayer.com/mcp/public

Where this comes from

  • Description: written from the project's website and its MCP Registry entry.
  • Status and tools: checked automatically by AIPoweredBot on 19 November 2026.
  • Install steps: from the official MCP Registry entry.

How AIPoweredBot checks servers