The MCP server and connector for your meetings.

Connect Kalima to Claude, ChatGPT, Cursor, or VS Code. Your assistant can search your transcripts, find what was said in a call months ago, and point you to the meeting it came from. Setup takes about two minutes, uses OAuth, and lets you choose additional summary and organization permissions.

Server URL

https://mcp.meetkalima.com/mcp

Set up ChatGPT Connect with Claude

See the 20 tools

Kalima MCP

What did we decide on the pilot budget?

Kalimasearch_recordings1 recording · Q3 Planning · 1:12:40

Kalimasearch_within_recording4 passages · 41:05 to 41:47

41:17 · Lea · “Finance approved it yesterday, so the money is there.”
41:28 · Marco · “Then we can start the pilot in April with the two teams in Hamburg.”
41:40 · Anna · “Marco, could you send me the approval from finance?”

Finance approved the forty thousand euros, so the money is there. Q3 Planning · 41:17 The pilot starts in April with the two teams in Hamburg, and Munich and Vienna follow in June if the first month goes well. Q3 Planning · 41:28 Finance wants a short report after the first month. Q3 Planning · 41:17

Who has to do what next?

Kalimaget_transcriptQ3 Planning · 38 segments

Anna updates the launch timeline today Q3 Planning · 0:23

Lea briefs the support team this week Q3 Planning · 0:23

Marco sends Anna the approval from finance Q3 Planning · 41:40

Where else did Munich come up?

Kalimasearch_all_transcripts3 matches in 2 recordings

In two other meetings. In the customer call on September 12, the Munich team asked for the workshops in German. Customer call · 18:04 In the team weekly, Lea suggested they start in June. Team weekly · 07:31

Draft a short note for the Munich team.

Hi everyone in Munich, a quick update from our Q3 planning: the pilot starts in April in Hamburg, and you follow in June if the first month goes well. As you asked, we will run the workshops in German. Any questions, just reply here.

Ask anythingAsk anythingAsk anythingAsk anythingAsk anythingWhat did we decide on the pilot budget?Who has to do what next?Where else did Munich come up?Draft a short note for the Munich team.Kalima

Add the connector once, then just ask.

There is nothing to install and no API key to manage. Your AI tool connects directly to Kalima through the Model Context Protocol.

01
Add the server URL
Paste the Kalima MCP URL into your AI tool's connector settings. The supported tools on this page can all connect to a remote MCP server.
02
Sign in and approve
Your browser opens a Kalima approval screen. Allow reading and optionally enable saved summaries or organizing recordings.
03
Ask about your meetings
Your assistant can now search transcripts, open a specific meeting, and show the exact passage it is quoting, all without leaving the chat.

Set the connector up in your tool of choice.

Step by step guides for the five AI tools people connect most. Whether your tool calls it a connector or an MCP server, the setup is the same and takes about two minutes.

Connect Kalima to ChatGPT

Add Kalima as a custom MCP server in ChatGPT. Enter the server address, then sign in to Kalima and approve access.

Requires: Developer mode enabled
Method: Custom MCP server

ChatGPTchatgpt.com

Kalima

ChatGPTchatgpt.com

Plugins

PublicPersonal

Search plugins

Create a pluginUpload a plugin archiveCreate a custom MCP server

Security and sign-in

Developer modeElevated risk

Create a custom MCP server

NameKalimaDescription (optional)Connection

Server URLTunnel

https://mcp.meetkalima.com/mcp

AuthenticationOAuth

Confirm to continueCreate as plugin

Authorize access

ChatGPTKalima

Recordings and transcripts

Allow selected access

Connected

Kalimahttps://mcp.meetkalima.com/mcp

ChatGPT setup

Turn on developer mode
Open Settings and choose Security and sign-in from the left sidebar, not Connectors. Scroll down to the Developer mode section and switch it on. ChatGPT marks it as elevated risk, because developer mode is what allows unverified connectors to be added at all.
Open the custom server dialog
Open Plugins in ChatGPT. Select Add at the top right, then Create a custom MCP server.
03
Enter Kalima's name and address
Enter Kalima as the name. Under Connection, choose Server URL and paste the address below, including /mcp. Keep authentication on OAuth. Confirm that you want to continue, then select Create as plugin.
Server URL
```
https://mcp.meetkalima.com/mcp
```
Approve access in Kalima
Sign in to Kalima if prompted. Review the permissions and select Allow selected access. You then return to ChatGPT.

Connect Kalima to Claude, on desktop and web

Claude supports remote MCP servers as custom connectors. Once added, the connector follows your account, so it works in Claude Desktop and on claude.ai alike.

Connect with Claude

Requires: Any Claude plan
Method: Custom connector

Open Settings, then Connectors
In Claude, open Settings and choose Connectors. You will see the connectors already on your account, alongside suggestions like Slack and Google Drive.
Open the Add menu, then Add custom connector
The Add button sits at the top right of the Connectors page. Open it and choose Add custom connector. Kalima is not in Claude's connector directory, so this is the way in.
03
Fill in the first two fields only
Enter Kalima as the Name, and paste the server URL into Remote MCP Server URL. Then press Add. That is the whole form.
Server URL
```
https://mcp.meetkalima.com/mcp
```
Leave Advanced settings closed. OAuth Client ID and OAuth Client Secret are both optional and should stay empty: Kalima registers your client automatically through dynamic client registration, so there is nothing to put in them.
Sign in and approve access
Claude opens Kalima in your browser. Sign in, review what the connector is asking for, and choose Allow selected access. You are returned to Claude with the connector enabled.
Claude reaches your connector from Anthropic's cloud rather than from your own machine. Kalima is a public endpoint, so this works without any tunnel or firewall change.
Ask your first question
Start a chat and ask something like: what did we decide in last week's planning meeting? Claude will call the Kalima tools and cite the meetings it drew from.

Connect Kalima to Claude Code

One command adds the server, and the OAuth flow runs from your terminal. Useful when you want meeting context while you are actually writing the code that came out of the meeting.

Requires: Claude Code CLI
Method: claude mcp add

01
Add the server
The --transport http flag is what marks this as a remote server rather than a local process.
Terminal
```
claude mcp add --transport http kalima https://mcp.meetkalima.com/mcp
```
02
Make it available everywhere
By default the server is scoped to the current project. Add --scope user to reach it from every project on your machine.
Terminal
```
claude mcp add --transport http kalima --scope user https://mcp.meetkalima.com/mcp
```
Authenticate
Run /mcp inside Claude Code and pick Kalima to start the sign-in, or run claude mcp login kalima straight from your shell. Either opens the Kalima consent screen in your browser.
04
Confirm it connected
The server should be listed as connected with its thirteen tools. The /mcp panel shows the same thing, plus a Re-authenticate option if a token ever needs refreshing.
Terminal
```
claude mcp list
```

Connect Kalima to Cursor

Cursor reads MCP servers from a JSON file, either globally for every project or checked into one repository. Remote servers need only a URL.

Requires: Cursor, any plan
Method: mcp.json

Open your MCP config
Use ~/.cursor/mcp.json to make Kalima available in every project, or .cursor/mcp.json in a repository to scope it to that project. Create the file if it does not exist yet.
02
Add the Kalima server
Cursor keys this map mcpServers, and a remote server needs nothing but a url. There is no transport or type field to set.
~/.cursor/mcp.json
```
{
  "mcpServers": {
    "kalima": {
      "url": "https://mcp.meetkalima.com/mcp"
    }
  }
}
```
Sign in
Open Cursor settings and find Kalima in the MCP list. It will show as needing login until you authenticate. Click through, approve read-only access in the browser, and the tools become available.
Use it in chat
In agent mode, ask about a meeting and Cursor will call the Kalima tools. Useful for turning a spec discussion straight into an implementation without re-reading the transcript yourself.

Connect Kalima to VS Code and GitHub Copilot

Copilot's agent mode can call MCP tools. Configure the server per workspace or globally, then use it from the chat panel.

Requires: VS Code with Copilot
Method: .vscode/mcp.json

01
Add the server config
Create .vscode/mcp.json in your workspace. Note that VS Code keys this map servers rather than mcpServers, and does require an explicit type.
.vscode/mcp.json
```
{
  "servers": {
    "kalima": {
      "type": "http",
      "url": "https://mcp.meetkalima.com/mcp"
    }
  }
}
```
Prefer a guided flow? Run MCP: Add Server from the Command Palette instead, and choose whether to install it for the workspace or globally.
Start the server
VS Code shows a Start action above the server entry in the JSON file. Run it, and sign in when the browser opens.
Switch Copilot to agent mode
Open the Chat view and select Agent from the mode picker. Tool calling is only available in agent mode, so the Kalima tools will not appear in ask mode.
Check the tools are there
Click the tools icon in the chat input to see everything Copilot can call. Kalima's thirteen tools should be listed and individually toggleable.

Using something else?

Kalima implements the Model Context Protocol over streamable HTTP with OAuth 2.1, so any spec-compliant client can add it as a connector using the same server URL. Windsurf, Zed, Cline, and Goose all work the same way: add the remote server, sign in, approve.

Sixteen tools, one connection.

Your assistant picks the right tool for the question. You never call these by hand.

Three built-in prompts

Ready-made flows your tool can offer as one-click actions, so common questions do not need to be typed out.

weekly_review

Walk through everything recorded in the past week.

catch_up_on_project

Get up to speed on one project's recent recordings.

find_mentions

Track down every time a topic, name, or account came up.

The answer is usually in a meeting you already had.

The information is usually there. Finding it is the hard part. Nobody wants to search through four hours of recordings for one sentence, so decisions get revisited and commitments slip through the cracks. Your assistant can find the relevant moment for you.

What objections came up most often in demos this quarter?
Summarize where the Meridian project stands based on our last four calls.
Did we ever promise that customer a custom SLA? Show me exactly what was said.
Draft the follow-up email from this morning's kickoff, using what we actually committed to.
Who has been in the most meetings about the migration, and what did each of them own?
Find the moment in Tuesday's review where we talked about the pricing tiers.

Access with your permission.

Giving an AI tool access to meeting data requires trust. Here is exactly what the Kalima connector can and cannot do.

You approve assignment changes
Reading is the default. Organizing requires a separate permission and confirmation of a preview. Moving to or from a shared project can change who has access. Transcripts and audio remain unchanged.
No API keys to create or paste
Kalima uses OAuth 2.1 with PKCE. You approve access in your browser, and your tool receives a limited token that expires automatically. Nothing sensitive needs to be stored in a configuration file.
It only sees meetings you can access
The connector follows your permissions in Kalima and narrows them further. Password-protected share links and open link-sharing grants are intentionally excluded.
Revoke access whenever you need to
Every connector you have approved appears in Settings, under Integrations, Connected apps. When you revoke one, access ends on its next request rather than waiting for the token to expire.
Separate from live transcription
Transcript requests have their own capacity limits. Organization requests are limited to 20 recordings per preview.
Saved summaries with your permission
If you allow it, your assistant can read compact excerpts from your saved AI summaries. This helps it suggest a project without loading the full transcript. It does not generate new summaries.

Kalima is operated in Germany and your data stays in the EU. We follow GDPR requirements and model our security controls on ISO 27001. We are not certified against it.

Included on every plan.

The Kalima connector works on every plan, the Free one included. How many requests you get per day and week depends on the plan.

Plan	Requests per day	Requests per week
Free	100	500
Plus	250	1,000
Pro	1,000	5,000
Team	5,000	25,000
Enterprise	Unlimited	Unlimited

One request is one tool call, not one question. Answering a single question often takes your assistant two or three calls, so budget accordingly.

Questions, answered.

The Model Context Protocol is an open standard for connecting AI assistants to outside data. An MCP server exposes a set of tools that an assistant can call on your behalf. The Kalima MCP server exposes sixteen tools over your meetings, so your assistant can search and read transcripts without you copying anything into the chat.

In practice, yes. Connector is the word most AI tools show you in their own interface: Claude and ChatGPT both call these Connectors, while Cursor and VS Code call them MCP servers. They are two names for the same thing, and Kalima works as both. Add the Kalima connector, or add the Kalima MCP server, and you get the same sixteen tools either way.

Anything that speaks the Model Context Protocol. This page has step by step guides for Claude Desktop, Claude Code, Cursor, ChatGPT, and VS Code with GitHub Copilot. Windsurf, Zed, Cline, and Goose connect the same way using the same server URL.

No. Kalima uses OAuth 2.1 with PKCE, so you sign in through your browser and approve access on a consent screen. Your tool receives a scoped token that expires and rotates on its own. There is no key to generate, paste, or rotate manually.

It cannot edit transcripts or delete recordings. If you grant organizing permission, it can change project and workspace assignments after you confirm a preview. Shared project access may change with the assignment.

Exactly the meetings you can already open yourself: the ones you own, ones shared with you directly, and ones you can reach through a project, team, or organization. Password-protected links and open link-sharing grants are excluded, so the MCP view is slightly narrower than the app.

All of them, including Free. The Free plan gets 100 requests a day and 500 a week. Limits rise through Plus, Pro, and Team, and Enterprise is unlimited. If you hit a limit, the connector tells you which one and when it resets.

Yes. Each connector authorizes separately and appears as its own entry in Settings, under Integrations, Connected apps, so you can run Claude and Cursor against the same account and revoke either one independently.

Open Settings, then Integrations, and revoke it under Connected apps. Access is cut on that connector's next request, not whenever its token happens to expire.

Yes, as a custom plugin. Turn on Developer mode under Settings, then Security and sign-in, then add Kalima from Plugins in the left sidebar. One detail matters: enter the full endpoint URL ending in /mcp. A bare domain will connect and then show no available actions.

Only if you separately allow it. The connector can then return compact excerpts from your own saved summaries, with a notice when freshness is uncertain. It does not generate summaries or charge for new AI generation.

Put your meetings where you already work.

Start recording with Kalima, add the connector to your AI tool, and stop rewatching calls to find one sentence.

Start for free See pricing

Tool	What it does
Organizing recordings
`list_workspaces`List your personal workspaces as destinations for recordings.	List your personal workspaces as destinations for recordings.
`get_recording_summaries`Read compact excerpts from saved summaries of your own recordings, with separate permission.	Read compact excerpts from saved summaries of your own recordings, with separate permission.
`move_recordings`Preview and confirm project or workspace assignments for up to 20 recordings.	Preview and confirm project or workspace assignments for up to 20 recordings.
`find_unassigned_recordings`Find up to 20 of your finished recordings that have no project or workspace yet, without changing anything.	Find up to 20 of your finished recordings that have no project or workspace yet, without changing anything.
`review_recording_assignments`Show suggested projects or workspaces for up to 20 recordings, which you edit and confirm before anything moves.	Show suggested projects or workspaces for up to 20 recordings, which you edit and confirm before anything moves.
`preview_recording_assignments`Check the chosen destinations in a read-only preview before you confirm a move.	Check the chosen destinations in a read-only preview before you confirm a move.
Finding a recording
`search_recordings`Full-text search across every recording you can access, ranked, with a snippet and timestamps.	Full-text search across every recording you can access, ranked, with a snippet and timestamps.
`list_recordings`Browse your recordings by recency, length, or name, filtered by project, team, status, or date range.	Browse your recordings by recency, length, or name, filtered by project, team, status, or date range.
`search_all_transcripts`Find the exact moments a topic came up across everything you recorded, with speaker and timestamp.	Find the exact moments a topic came up across everything you recorded, with speaker and timestamp.
Reading transcripts
`get_transcript`Paginated transcript for one recording. Ask for full detail, speaker labels only, or plain text.	Paginated transcript for one recording. Ask for full detail, speaker labels only, or plain text.
`show_transcript`Opens a recording's transcript as a view right in the chat, with speakers, timestamps and a link into Kalima. Text-only assistants get the words instead.	Opens a recording's transcript as a view right in the chat, with speakers, timestamps and a link into Kalima. Text-only assistants get the words instead.
`get_live_transcript`Follow a recording while it is still running: only what has been said since your assistant last looked, plus corrections to what it already had.	Follow a recording while it is still running: only what has been said since your assistant last looked, plus corrections to what it already had.
`search_within_recording`Jump to where a topic came up inside a single recording instead of reading the whole thing.	Jump to where a topic came up inside a single recording instead of reading the whole thing.
Recording detail
`get_recording`One recording's metadata: status, mode, start, duration, participants, word count.	One recording's metadata: status, mode, start, duration, participants, word count.
Projects
`list_projects`Your project folders and their sub-projects, with recording counts and last activity. Searchable by name.	Your project folders and their sub-projects, with recording counts and last activity. Searchable by name.
`get_project`One project in full, including total recorded time, its place in the folder tree, and its most recent recordings.	One project in full, including total recorded time, its place in the folder tree, and its most recent recordings.
People and patterns
`list_speakers`Everyone you have named in Kalima, with how often and how recently they spoke.	Everyone you have named in Kalima, with how often and how recently they spoke.
`get_recording_stats`Totals across your account: recordings, recorded hours, words transcribed, first and last one.	Totals across your account: recordings, recorded hours, words transcribed, first and last one.
Workspace
`list_organizations`The organizations you belong to and your role in each.	The organizations you belong to and your role in each.
`list_teams`The teams you belong to, with role and parent organization.	The teams you belong to, with role and parent organization.

The MCP server and connector for your meetings.

Add the connector once, then just ask.

Add the server URL

Sign in and approve

Ask about your meetings

Set the connector up in your tool of choice.

Connect Kalima to ChatGPT

Turn on developer mode

Open the custom server dialog

Enter Kalima's name and address

Approve access in Kalima

Connect Kalima to Claude, on desktop and web

Open Settings, then Connectors

Open the Add menu, then Add custom connector

Fill in the first two fields only

Sign in and approve access

Ask your first question

Connect Kalima to Claude Code

Add the server

Make it available everywhere

Authenticate

Confirm it connected

Connect Kalima to Cursor

Open your MCP config

Add the Kalima server

Sign in

Use it in chat

Connect Kalima to VS Code and GitHub Copilot

Add the server config

Start the server

Switch Copilot to agent mode

Check the tools are there

Using something else?

Sixteen tools, one connection.

Three built-in prompts

The answer is usually in a meeting you already had.

Access with your permission.

You approve assignment changes

No API keys to create or paste

It only sees meetings you can access

Revoke access whenever you need to

Separate from live transcription

Saved summaries with your permission

Included on every plan.

Questions, answered.

Put your meetings where you already work.