Connect Perplexity, Gemini, Cursor & Other Clients

Connect Tenali to Perplexity, Gemini, Cursor, VS Code, GitHub Copilot, Windsurf and other supported MCP clients using the same Model Context Protocol server URL.

Tenali works with MCP clients that support remote servers over streamable HTTP with OAuth — which covers most of them. Clients that only accept local or API-key-based servers can't connect. Perplexity and Gemini follow the same custom-connector pattern as Claude and ChatGPT; Cursor and VS Code install in one click.

Every client uses the same server URL: https://api.tenali.ai/api/mcp.

Perplexity

Available on Perplexity Pro and Max.

1
Open connector settings
In Perplexity, open your Account → Settings → Connectors.
2
Add a custom connector
Click "+ Custom connector" in the top-right and choose Remote.
3
Paste the server URL
Paste https://api.tenali.ai/api/mcp and click Add.
4
Sign in and confirm
Click the new connector card and sign in with your Tenali account. Ask about a recent meeting to verify it works.

Gemini

Gemini support is verified as of 2026-07-26. This is the fastest-moving eligibility surface on this page — US Google AI Pro subscribers only gained Gemini Spark on 2026-07-16 — so re-check the linked Google pages if you're reading this much later.

Gemini supports MCP servers as custom Connected Apps for Gemini Spark. Eligibility is narrow — check all of these before you start:

A personal Google Account. Work and school accounts aren't supported.
Age 18 or over, and located in the US.
A qualifying Google AI Pro or Ultra subscription — this is what grants access to Gemini Spark. Either tier qualifies, since custom apps are US-only.
Keep Activity turned on — formerly called Gemini Apps Activity. Gemini can't connect to custom apps while it's off.

1
Open Connected Apps
At gemini.google.com, click Settings & help → Connected Apps. If you don't see it, click Personal Intelligence → Connected Apps first.
2
Add a custom app
Under Custom apps for Spark, paste https://api.tenali.ai/api/mcp into the add-a-custom-app field and click Next.
3
Authorize and confirm
Sign in with your Tenali account to authorize it, then ask about a recent meeting to verify.

Cursor and VS Code

Both support one-click install straight from Tenali.

1
Open the AI Assistants tab
In Tenali, go to Settings → AI Assistants and scroll to Developer tools.
2
Click the install button
Click Add to Cursor or Install in VS Code. Your editor opens and pre-fills the Tenali MCP server for you.
3
Sign in and confirm
Approve the install in your editor, then sign in with your Tenali account when prompted. Ask about a recent meeting to verify.

Any Other MCP Client

Windsurf, Zed, Goose, Raycast and other clients that support remote MCP servers with OAuth work too. The pattern is always the same:

Find where your client adds a remote or custom MCP server.
Paste https://api.tenali.ai/api/mcp as the server URL.
Choose OAuth if asked how to authenticate.
Sign in with your Tenali account and approve the permissions.

GitHub Copilot

Copilot support depends on which surface you're using, because Tenali authenticates with OAuth and has no API-key option.

Works — Copilot in VS Code, Visual Studio, and the Copilot CLI. All three support remote MCP servers with OAuth.
Doesn't work — Copilot's coding agent and code review. GitHub's documentation states these don't support remote MCP servers that use OAuth.
Untested — JetBrains, Xcode and Eclipse. GitHub added OAuth support for these in preview, so they may well work, but we haven't verified the flow against Tenali yet. Try it if you like; we'll list it here once we've confirmed it.

Troubleshooting

The client won't authenticate — make sure it supports OAuth-based remote MCP servers. Clients that only accept API-key servers can't connect to Tenali.
"Connection request expired" — the sign-in handshake lasts ten minutes. Start the connection again from your client.
The AI Assistants tab isn't in Tenali — the MCP server is enabled per workspace. Contact your Tenali admin or security@tenali.ai.