# TagoIO MCP Server v3.0.0 — Remote HTTP Support

**URL:** <https://community.tago.io/t/tagoio-mcp-server-v3-0-0-remote-http-support/1887>\
**Category:** Announcements\
**Created:** [April 14, 2026, 9:32pm UTC](https://community.tago.io/t/tagoio-mcp-server-v3-0-0-remote-http-support/1887 "2026-04-14T21:32:22Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![vitorfdl](https://yyz2.discourse-cdn.com/flex034/user_avatar/community.tago.io/vitorfdl/32/1728_2.png) [@vitorfdl](https://community.tago.io/u/vitorfdl)\
**Post date:** [April 14, 2026, 9:32pm UTC](https://community.tago.io/t/tagoio-mcp-server-v3-0-0-remote-http-support/1887/1 "2026-04-14T21:32:22Z")

</div>

The TagoIO MCP Server connects your AI assistant directly to your TagoIO account. Ask it to check device data, generate Analysis scripts, debug issues, or build reports — and it works with the actual context of your devices, dashboards, and configurations.

Here’s what it can do:

- **Device management** — List devices, read configurations, inspect real-time data.
- **Data analysis** — Run statistical operations on stored data: sums, averages, custom reports.
- **Code generation** — Generate TagoIO Analysis scripts with proper platform context and imports.
- **Platform access** — Retrieve users, actions, analysis scripts, and account statistics.
- **Debug support** — Trace issues and analyze tag relationships across your account.

### What’s new in v3.0.0

**Remote server.** We’re hosting an MCP server at `https://mcp.ai.tago.io`. Point your IDE or AI tool to that URL with your Profile Token, and you’re connected. The [setup guide](https://github.com/tago-io/mcp-server/) has ready-to-copy configurations for 15+ platforms — VS Code, Claude Code, Cursor, Windsurf, JetBrains, GitHub Copilot, Gemini CLI, and more.

**TagoDeploy dedicated instances.** If you run a dedicated instance, add a single `x-tagoio-region` header with your instance URL and every request routes to your private environment. Works with both the hosted server and self-hosted setups.

**Self-hosted HTTP transport.** Prefer to run it yourself? The same HTTP Streamable transport is available locally via `npx -y @tago-io/mcp-server http`, or you can deploy it to AWS Lambda for a serverless setup.

STDIO transport and all existing tools continue to work the same way.

### Get started

Grab your Profile Token from [Profile Settings](https://admin.tago.io/profile), pick your platform from the [setup guide](https://github.com/tago-io/mcp-server/), and connect in a couple of minutes.

Questions or feedback — drop them here.