[Skip to content](https://www.innguma.com/en/resources/faq/compatible-assistants/#contenido)

Frequently asked questions

# Which AI assistants are compatible with Innguma?

**Short answer** 

Four: ChatGPT, Claude, Microsoft Copilot and Gemini. Innguma offers connectivity through the Model Context Protocol (MCP), currently compatible with these four assistants. Through MCP, they work with the knowledge and capabilities that Innguma makes available to each integration. Each provider decides in which products and plans MCP connectors can be added, and who can activate them.

## Why does an open standard not mean ‘any assistant’?

MCP is an open standard: any application can implement it[\[1\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-1). But each assistant implements a different part of the protocol, with its own requirements for connection, authentication and confirmation. Microsoft, for example, indicates that Copilot Studio currently supports MCP tools and resources, not templates[\[2\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-2); Google only supports the Streamable HTTP transport in Gemini Enterprise[\[3\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-3). That is why Innguma only describes as compatible the assistants with which its connectivity has been validated: ChatGPT, Claude, Microsoft Copilot and Gemini.

## How is an MCP server connected in each assistant?

A summary of each provider’s official documentation, consulted in October 2026. These conditions change often: check them at source before deciding.

| Assistant | Where the connector is added | Who activates it | Actions that modify data |
| --- | --- | --- | --- |
| ChatGPT | Developer mode and MCP apps, in the web version | Administrators or owners of the workspace (Business, Enterprise, Edu) | Full support, including write actions, is in beta in Business, Enterprise and Edu; ChatGPT may ask for confirmation depending on context[\[4\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-4) |
| Claude | Custom connectors with remote MCP (web, desktop and Cowork) | On Team and Enterprise plans, owners; then each member connects | Anthropic advises turning off write tools when using research mode, which invokes them without additional approval[\[5\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-5) |
| Microsoft Copilot | Copilot Studio, with the assistant for adding MCP servers | Whoever creates or manages the agent; requires generative orchestration | Microsoft reminds that whoever connects an external server answers for the tools it uses[\[2\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-2) |
| Gemini | Gemini Enterprise, as a data store for a custom MCP server | An administrator with the Discovery Engine editing role | Tools appear as actions disabled by default and, by default, each call asks the user for confirmation[\[3\]](https://www.innguma.com/en/resources/faq/compatible-assistants/#ref-3) |

## What can the assistant do once connected?

With the capabilities that your organisation enables in its integration, from any of the four assistants you can run the watch cycle in Innguma: consult and request summaries or comparisons of the news items in the inbox, organise and add sources, modify filters, audit the performance of sources, publish content, edit or delete what has been published, and search across it. What is automated, and what goes through human review first, is decided by your team.

The assistant you choose does not change what Innguma makes available to the integration. It changes the experience: how it asks for confirmations, how it shows the tools, and which data conditions it applies.

## What should you check before connecting your assistant?

1.  **Plan and edition.** Verify that your organisation’s plan supports remote MCP connectors and, if you are going to use write capabilities, that it supports those too.
2.  **Who administers.** Identify who can add the connector in the assistant: on the business plans of all four assistants this is an administrator profile.
3.  **Internal AI policy.** Confirm which assistants your organisation authorises for corporate information.
4.  **Data handling.** What reaches the assistant is governed by that provider’s terms. Review them with your systems and data protection teams.
5.  **Confirmations.** Check how each assistant asks for approval before executing an action, and whether you can disable specific tools.

## Sources and references

1.  \[1\] [Model Context Protocol. Specification (revision 2026-07-28)](https://modelcontextprotocol.io/specification/latest)
2.  \[2\] [Microsoft Learn. Extend your agent with Model Context Protocol (Copilot Studio). Updated in 2026](https://learn.microsoft.com/en-us/microsoft-copilot-studio/agent-extend-action-mcp)
3.  \[3\] [Google Cloud. Set up your custom MCP server data store (Gemini Enterprise). Accessed October 2026](https://docs.cloud.google.com/gemini/enterprise/docs/connectors/custom-mcp-server/set-up-custom-mcp-server)
4.  \[4\] [OpenAI Help Center. Developer mode and MCP apps in ChatGPT. Accessed October 2026](https://help.openai.com/en/articles/12584461)
5.  \[5\] [Claude Help Center. Get started with custom connectors using remote MCP. Accessed October 2026](https://support.claude.com/en/articles/11175166)

## Related

- [What is MCP (Model Context Protocol)?](https://www.innguma.com/en/resources/faq/what-is-mcp/)
- [Is MCP available in Innguma?](https://www.innguma.com/en/resources/faq/mcp-available/)
- [Can AI publish or change things on its own?](https://www.innguma.com/en/resources/faq/ai-publishing/)
- [Academy](https://academy.innguma.com/en/)

- [Request a demo](https://www.innguma.com/en/demo/)
- [Back to the FAQ](https://www.innguma.com/en/resources/faq/#faq-ia)