LabGiant,

LabGiant in Claude and ChatGPT

Tell your AI what you need done. It finds the core facility on LabGiant that does it, fills in that facility's request from your words, and sends it when you say yes.

What it does

  • Finds the services, instruments, facilities and experts on LabGiant that match what you describe.
  • Opens a facility the way its own assistant reads it: the services you can reach, the facility's line on when to pick each one, and the answers it has written, quoted as the facility's own words.
  • Reads what the facility asks before it takes on a request, and the limits it sets.
  • Checks your request before anything is sent, with LabGiant's estimate at the rate that applies to you. The facility's team confirms the quote.
  • Sends the request in your name once you say yes, and tells you what is still to answer or attach on LabGiant.

Add it to Claude

  1. In Claude, open Settings, then Connectors.
  2. Find LabGiant in the directory and press Connect. If it isn't listed for you yet, choose Add custom connector, name it LabGiant, and use this address: https://api.labgiant.com/mcp
  3. Start a new chat and describe the work you need done.

Claude reads LabGiant without signing in. The first time you send a request, it asks you to sign in to LabGiant and press Allow.

Add it to ChatGPT

  1. In ChatGPT, open Plugins and search for LabGiant, then add it.
  2. If it isn't listed for you yet, add a custom MCP server instead. Name it LabGiant, use this address: https://api.labgiant.com/mcp and choose OAuth as the sign-in, with the client ID and secret left empty.
  3. Sign in to LabGiant and press Allow.
  4. In a new chat, ask for LabGiant and describe the work you need done.

OpenAI decides which ChatGPT plans can send through a custom server. On some plans it can read but not send.

When you send

Your AI shows you everything that goes first: your project in your own words, and the answers to the facility's questions. Claude and ChatGPT both ask you before the send runs, and nothing goes until you say yes. The facility receives it the way it receives a request from its own page on LabGiant, and you get a receipt by email with a link to the request.

Files the facility asks for are attached on LabGiant after you send, from the link in the receipt.

What it can and can't do

It can

  • Search LabGiant and open any listing or facility you could see on the page.
  • Read a facility's request form, its limits and its written answers.
  • Check a request and give LabGiant's estimate for it.
  • Send a request for a service, or ask for a staff run or training on an instrument.

It can't

  • Book instrument time.
  • Change or cancel a request, or reply on its thread. Do that on LabGiant.
  • Attach files or take payments.
  • See anything LabGiant wouldn't show you on the page.
  • Speak for a facility beyond what the facility wrote. Anything your AI adds is its own suggestion, never the facility's answer.

Your data

LabGiant receives what your AI sends with each step: what you search for, and for a request, your description and your answers. It never reads the rest of your conversation.

Signing in gives Claude or ChatGPT a key that opens the connector and nothing else. With it, your AI can find facilities, check a request with your labs and your rate, and send requests in your name. It can't open anything else in your account. Removing LabGiant from your AI app stops it from using LabGiant, and Disconnect under Your AI assistant, in your LabGiant account settings, ends the key. Our privacy policy has the rest.

Help

Something not working, or a facility you expected to find? Write to [email protected] and we'll answer you.