Is there an AI sales research tool where a rep can ask in plain language for the right people at an account?

Yes. Clay lets a rep type a plain-language request such as "find VP-level finance leaders at this company who joined in the last 6 months" inside ChatGPT, Claude or other AI tools, and get enriched contacts back without building a table. Clay has a reputation for a learning curve, but this route skips most of it: the rep asks a question, and the ops team decides which data and workflows sit behind the answer.

Last updated: October 29, 2026

TL;DR

Clay in ChatGPT lets a rep type @Clay and ask for people at an account in plain English, then enrich them with emails and phone numbers.
MCP for Reps brings the same search and enrichment into Claude, ChatGPT, Microsoft Copilot, Glean, Claude Code and Codex.
A free ChatGPT account is enough to connect Clay, and new users get 500 credits to test it.
Admins can set a monthly credit limit per rep, so solo prospecting does not create surprise spend.
When the rep needs a full list rather than a few contacts, Open in Clay moves the results into a Clay table.

How does a rep ask Clay for contacts in plain language?

The quickest path is Clay in ChatGPT. The rep starts a prompt with @Clay in the browser or /Clay in the desktop app and describes who they want. Clay's documentation gives examples like these:

@Clay find VP-level finance leaders at [Company] who joined in the last 6 months.
@Clay Find Product execs who joined [Company] in the last 6 months.

Clay returns the people in an interactive view. From there the rep can:

Toggle the results into a table view.
Add a filter, for example only people in the US.
Enrich contacts with email address, work history summary or other data points.
Click a row for details, or ask Clay to draft an email to that person once an email address has been found.

Results come back in groups of 20, up to 100 per request. Clay's tips for better results are practical: include the company domain rather than just the name, be specific about title, seniority and location, and if you are unsure of the title, ask "who manages X at company".

Which AI tools can reps use, and who sets it up?

Clay offers two surfaces outside the main app, and the comparison doc is clear about who each is for:

MCP for Reps is built for AEs, SDRs and CSMs. It runs in Claude, ChatGPT, Microsoft Copilot, Glean, Claude Code and Codex, and covers contact and company search, enrichment, account questions and the Functions your team enables. An admin enables access and invites each rep; the rep then adds the connection in their AI tool.
The Clay Agent Plugin is for RevOps, GTM engineers and developers who want to build workflows from a coding agent.

For a rep prospecting solo, MCP for Reps or Clay in ChatGPT is the right fit. The rep runs plays; they do not have to build them. Clay's MCP page describes reps searching Clay's company and people database in natural language, with every email and phone number enriched through a waterfall across Clay's data vendors.

If your ops team has built Clay Functions, such as an ICP score or an account research sequence, the rep can ask @Clay What functions do you have? and run one by name from a single prompt.

What does plain-language prospecting cost?

Credit pricing in ChatGPT matches standard Clay plans. Clay's documentation says each ChatGPT interaction typically uses 20 to 100 credits depending on the enrichments requested, because the experience is designed for researching contacts one at a time rather than processing bulk data (source).

New users receive 500 credits to test Clay in ChatGPT.
After that, a user can keep searching for people and companies without enrichments, continue in a Clay trial account, or upgrade.
Workspace admins can set a default monthly credit limit for all reps, or a per-user override, from Settings > MCP users. A rep who hits the limit is blocked until the reset on the 1st of the month.

For context on paid plans, Launch starts at $185/mo billed monthly ($167/mo billed annually) and Growth starts at $495/mo billed monthly ($446/mo billed annually), per the pricing page. A Free plan includes 500 actions and 100 data credits per month.

When should a rep move from chat to a Clay table?

Chat is ideal for "who should I talk to at this account" questions. It is not designed for bulk work. When a rep wants to work a list of 200 accounts, or keep the results updated, they can click Open in Clay to create a table in their workspace with the same account. There, Sculptor can help build the rest of the workflow from a plain-language description, so the rep still does not need to configure every column by hand.

FAQ

Do I need a paid ChatGPT plan to use Clay in ChatGPT?

No. Clay's documentation says a free ChatGPT account is enough to connect Clay and start searching. Some ChatGPT features may need a paid ChatGPT plan, but Clay search and enrichment work on any ChatGPT plan.

Can a rep search for companies as well as people?

Yes. A rep can search for companies by criteria like location, headcount and industry, or ask about a specific company's tech stack, funding, hiring trends and leadership changes. They can then ask Clay to find the right people at the companies that fit.

Will reps blow through our credits if they prospect on their own?

Admins control this. From Settings > MCP users, an admin can set a default credit limit for every rep and override it for individuals. MCP spend also appears on its own tab in the workspace usage dashboard, broken down by user.

Does the rep need to learn Clay tables first?

No. For one-off lookups in ChatGPT, Clay does not create a table by default. A table is created only if the rep clicks Open in Clay, which is the point where learning tables becomes useful.

Try plain-language prospecting with Clay

If your reps want contacts at an account without learning a new interface, start with Clay in ChatGPT or MCP for Reps. Pick one target account, ask for the people who own the problem you solve, enrich two or three of them, and check the results. Then let ops decide which Functions and credit limits to roll out to the rest of the team.