Skip to main content

Intento

Intento MCP server

Company-approved translation for AI assistants and agents

Translate text and documents in the assistants your teams already use, and add translation to your AI agent workflows. Intento connects both to the Intento Enterprise Language Hub, with the same company requirements as your website and help desk. Intento AI Agents check terminology, tone of voice, and formatting and repair errors they find.

650+ languages and dialects · Unlimited seats

Trusted by 60+ global enterprises for translation automation.

IndeedHexagonStravaAgilentGovernment of Canada

650+ languages and dialects

Translate text and documents in AI assistants and AI agents.

Company requirements

Intento AI Agents check translations against your terminology, tone of voice, and formatting requirements and repair errors they find.

Any AI vendor

Combine models from multiple vendors and company-hosted deployments under one Intento contract.

Unlimited seats

Give every employee access. You pay for translation usage.

Apply terminology, tone of voice, and formatting rules in every language

Intento AI Agents check translations against your requirements and repair errors they find before returning the result. For example, they check that a product name stays unchanged and that a support reply uses the required formal tone.

For each language, Intento Language Engineers build and evaluate AI solutions on your content, including agentic workflows with Intento AI Agents, and choose the simplest one that meets your requirements.

Choose the simplest workflow that meets your requirements. In our State of Translation Automation 2026, a one-step solution with prompt and context engineering met Standard requirements in 38 of 100 languages. These requirements cover terminology, tone of voice, and formatting. Other languages needed more complex workflows.

Manage translation requirements centrally across assistants, agents, and business systems

Your localization team maintains company requirements such as terminology in Intento Language Hub. Each use case can add its own requirements, for example tone of voice, with AI solutions optimized for it.

Translation in AI assistants and AI agents meets the same company requirements as your website, help desk, TMS, Translation Portal, and Office add-ins. Update a glossary in Intento Console, and connected assistants, agents, and systems use the update through the translation workflows configured with that glossary.

Intento also reuses approved translations from Translation Storage in every assistant and agent. When the company adds or changes an AI assistant, employees connect the same server and keep the same company AI solutions.

Give employees, developers, and AI agents company-approved translation

Employees in AI assistants

Support, sales, marketing, legal, and HR teams translate text and documents in Claude, ChatGPT, and Copilot. Each employee signs in once with an Intento account.

Developers and AI agents

Developers translate in Claude Code, GitHub Copilot, and Cursor. Your AI agents call the same translation tools in their workflows and connect with an API key.

Localization and IT teams

The localization team maintains company requirements in Intento Language Hub. IT approves one connector for each assistant and sees usage per user in Intento Console.

Translate text and documents for every team and workflow

Employees ask for a translation in plain words, and AI agents call the Intento translation tools in their workflows. Word, Excel, PowerPoint, text, HTML, XML, and subtitle files return in the same format with the original layout, and PDFs and images come back as editable Word documents.

AI assistant
Employee
Translate this reply into Brazilian Portuguese.
Intento Translate

Customer support

Replies in the customer language, with the approved product names and support terminology.

AI assistant
Employee
Translate the attached product brief into German and Japanese.
Intento Translate

Sales and marketing

Documents in the original layout, with the brand terminology and tone of voice.

AI assistant
Employee
Translate this policy update into French.
Intento Translate

Legal and HR

Company documents under company data terms, in the formal tone of company communications.

AI assistant
Employee
Translate the new UI strings in this file into Spanish.
Intento Translate

Developers

Product terminology from the localization program, in Claude Code, GitHub Copilot, and Cursor.

AI agent
Tool call
translate · target languages: de, ja, pt-BR
Intento Translate

AI agents

Multilingual output in agent workflows, from the same company AI solutions.

Example requests are illustrative.

Connect AI assistants and agents through the Model Context Protocol

The Model Context Protocol (MCP) is an open standard for connecting AI applications to tools and data. Claude, ChatGPT, Microsoft Copilot Studio, GitHub Copilot, and Cursor support it, and so do many AI agent frameworks.

Through the Intento MCP server, we offer translation as MCP tools: text translation, document translation in the original layout, and a list of the company AI solutions available to the account. An AI assistant or agent calls these tools during a task and gets the translation back from Intento Language Hub.

Read more about MCP, APIs, and the A2A protocol in our blog post.

Use approved translation in your own AI agents

Connect custom AI agents, such as Copilot Studio agents or agents on an agent framework, and third-party AI agents with MCP support. Each agent calls the Intento MCP server like any other tool, with an Intento account or an API key.

Translation in agent workflows meets the same company requirements as your website, help desk, and TMS, with AI solutions optimized for each workflow. With a separate API key for each agent, you see the usage of each agent in Intento Console.

For batch translation and build pipelines, use the Intento API.

{
  "mcpServers": {
    "intento": {
      "url": "https://mcp.inten.to/v1/mcp",
      "headers": { "apikey": "YOUR_INTENTO_API_KEY" }
    }
  }
}

MCP configuration with an API key. For the EU data center, use https://mcp.eu.inten.to/v1/mcp.

Choose the AI models used for translation

Use your preferred AI vendor, or combine models from multiple vendors and company-hosted deployments in your company AI solutions, under one Intento contract.

When we change the model behind a solution, every connected assistant and agent gets the new version with the same settings. Employees can also name a specific engine in a request, if your administrators allow engine choice.

Pay for translation usage, not seats

Give employees access without an Intento seat charge for each person. You pay for translation usage. Setup and rollout depend on the assistant.

If you own the assistant rollout, you bring company translation rules and data terms to employees through the approved connection for each assistant.

Set up Claude, ChatGPT, Copilot, and AI agents

You need an active Intento contract and an account configured for translation.

1Add the server.

Add the Intento MCP server (Intento Translate) in the connector settings of the assistant, or in the MCP configuration of the agent.

2Sign in.

Sign in with an Intento account, or add an API key for AI agents and other MCP clients.

3Ask for a translation.

Employees ask in plain words in the chat, and agents call the translation tools in their workflows.

Claude

Claude on the web, desktop, and mobile, as a custom connector. On Team and Enterprise plans, an Owner adds it for the whole organization, and teams send every translation request in a shared Claude Project to Intento with one project instruction.

ChatGPT

ChatGPT, as a custom MCP server with Intento sign-in. Add a standing instruction in custom instructions or a ChatGPT Project to send translation requests to Intento.

Microsoft Copilot

Microsoft 365 Copilot, through a Copilot Studio agent with the Intento MCP server as a tool. For SharePoint, Intento has a document translation agent in Copilot Studio and configures it for your sites, with the translation next to the original in the same folder.

Developer tools

Claude Code and GitHub Copilot in VS Code, with Intento sign-in, and Cursor, with an API key.

Any MCP client or AI agent

Connect any MCP client with Streamable HTTP, including custom and third-party AI agents, with an Intento account or an API key.

US https://mcp.inten.to/v1/mcp
EU https://mcp.eu.inten.to/v1/mcp
User guide
Assistant or clientConnectionSign-inRollout
Claude (web, desktop, mobile)Custom connectorIntento accountOwner adds it for the organization on Team and Enterprise plans
Claude CodeMCP server in the command lineIntento account or API keyEach developer
ChatGPTCustom MCP server in developer modeIntento accountEach user
Microsoft 365 CopilotCopilot Studio agentAPI key in the agentIT deploys the agent
GitHub Copilot in VS CodeMCP configuration in VS CodeIntento accountEach developer
Cursor and other MCP clientsMCP configurationAPI keyEach client
AI agentsMCP configuration of the agentAPI key or Intento accountEach agent

Follow the steps for each assistant in the Intento MCP server user guide.

Leave server hosting and maintenance to Intento

We host the Intento MCP server in the US, the EU, or a dedicated environment, and maintain it, including fixes and updates. Because we host it, employees can use it in Claude on the web, desktop, and mobile, and in ChatGPT.

Get expert help from Intento Language Engineers

Our Language Engineers can identify your language requirements, configure the AI workflows behind every connected assistant and agent, and keep them optimized as models change.

They also configure assistant integrations, such as the SharePoint document translation agent in Copilot Studio.

Review access control, data handling, and certifications

Access control

  • Approval of the connector in each assistant by its administrators
  • Tool permissions in Claude for the whole organization: Always allow, Needs approval, or Blocked
  • Sign-in with an Intento account for each employee (OAuth 2.1), or an API key for AI agents and other MCP clients
  • Administrator control of the company AI solutions and engines for each account
  • Usage per user and per API key in Intento Console

Data handling

  • Encrypted storage of document jobs only, for 30 days by default
  • TLS 1.2 or higher in transit and AES-256 at rest
  • US, EU, or dedicated data center
  • “No share, no use, no trace” data terms with AI vendors

Certifications

  • ISO/IEC 27001:2022 certification for Intento Language Hub
  • ISO 9001:2015 certification for Intento services

Find answers on MCP, billing, and data security

MCP, the Model Context Protocol, is an open standard for connecting AI applications to tools and data. With the Intento MCP server, AI assistants and agents use translation from Intento Language Hub as an MCP tool.

You can connect Claude on the web, desktop, and mobile, Claude Code, ChatGPT, GitHub Copilot in VS Code, and Cursor directly, and Microsoft 365 Copilot through a Copilot Studio agent. Any MCP client with Streamable HTTP, including custom and third-party AI agents, connects with an API key or an Intento account.

Companies pay for translation usage, with unlimited seats for employees and AI agents. Contact our team for pricing.

We store document jobs only, encrypted, for 30 days by default, and process translations in the US, the EU, or a dedicated data center. Intento Language Hub has ISO/IEC 27001:2022 certification, and we work with AI vendors under “no share, no use, no trace” data terms.

Use MCP when an employee or an AI agent asks for a translation during a task. Use the Intento API for batch translation, build pipelines, and system integrations.

Yes. Each employee signs in with an Intento account, and you see usage per user in Intento Console. AI agents and other MCP clients connect with an API key.

Word, Excel, PowerPoint, text, HTML, XML, and subtitle files return in the same format, and PDFs and images go through OCR and come back as editable Word documents. Translate up to 100,000 characters of text per request. The file limit is 25 MiB; documents of a few pages work best. Use the Translation Portal or the Intento connector in your TMS for longer documents and large batches.

Yes. The connector uses the company AI solutions on Intento Language Hub, with the same company requirements as your TMS, website, and help desk, and each use case can add its own.

Read more on MCP and AI translation

Blog post

How to understand the difference between API, MCP, and A2A in enterprise localization

MCP compared with APIs and agent-to-agent protocols.

Report

State of Translation Automation 2026

AI solutions and company requirements in 100 target languages.

Product

Intento Translation Portal

Company-approved AI translation for every employee, in one portal.

Bring company translation rules into your AI assistants and agents

See how Intento can apply your terminology, tone of voice, and formatting requirements in your assistants and agent workflows.

Product names and logos belong to their owners. They identify customers, AI assistants, and AI model providers and do not imply endorsement.