B2B leads for startups, from the agent you already use
Leadsets lets founders and small sales teams ask Claude, ChatGPT or any MCP client for their next 100 prospects. It returns people, companies and verified work emails, each with match reasons and checked dates. It starts free with 1,000 enrichment records and 100 search records, no card. Paid plans are flat monthly prices from $29 with no credits, shared by the whole team. It has no phone numbers yet.
| Job | What you need | How Leadsets does it |
|---|---|---|
| Find the next 100 prospects | People who match your ICP, without learning a dashboard | search_companies and search_people from a plain-language prompt, with match reasons on every result |
| Get emails you can send to | Addresses that will not bounce | enrich_person returns a verified work email when one exists; verify_email labels verified, catch-all, unknown or invalid |
| Time your outreach | A reason to write this week | scrape_jobs reads careers pages for open roles with dates (scrape units, paid plans) |
| Research an account before a call | What they sell and to whom | scrape_company turns a homepage into products, positioning and customers named |
| Load the sequencer and CRM | A clean handoff | Skills and CSV for Instantly and Smartlead, push to LinkedCamp, skills and API for HubSpot, Pipedrive and Attio |
| Keep costs predictable | No credit math | Flat plans from $29/mo, records count only when data comes back, get_usage is free |
Where data gets in the way for small teams
Without a RevOps person, the founder or first seller does every go-to-market job: deciding who to sell to, building the list, finding emails, checking they will not bounce, researching accounts before a call, loading a sequencer and keeping the CRM tidy. The data part is usually where the hours go.
- Credits to count. When every lookup draws on a credit balance, building a list turns into a budgeting exercise.
- Emails you cannot trust. Stale titles and unverified addresses cost more in bounces and domain reputation than they save in time.
- Per-seat pricing. Adding a cofounder or a first SDR should not mean buying another licence.
- Too many vendors. One tool for search, another for verification and a third for company research means three keys, three bills and three exports to stitch together.
If you are still working out who to target, start with how to build a prospect list.
How Leadsets fits each job
Leadsets is a B2B data API and remote MCP server. Add https://mcp.leadsets.ai to Claude, ChatGPT, Cursor or any MCP client. Nothing is installed. Then describe your customer the way you would brief an SDR, for example "RevOps leads at 50 to 200 person fintechs". The agent turns that into filters and picks from eight tools. Setup steps for each client are in the MCP docs.
search_peoplefinds people by title, function, seniority, location, company size and industry.search_companiesfinds accounts by industry, headcount band, technology and location. Every result says why it matched.enrich_personreturns current role, company and a verified work email when one exists.enrich_companyreturns firmographics for a domain, with source and freshness on every field.verify_emaillabels each address verified, catch-all, unknown or invalid, with a confidence score. Uncertain results are never rounded up to valid, and invalid emails are never served as contacts.scrape_companyandscrape_jobsturn a homepage or careers page into structured facts: products, positioning, customers named, and open roles with dates.get_usagetells the agent what is left in the current window and week, at no cost.
Paid plans are a flat monthly price with no credits and no per-record bill. Records count only when data comes back, so an empty search uses no records. Emails returned by enrich_person are already verified, so there is no second tool and no second bill. Leadsets allows one account per company, and team members join that same account.
Two things help a team with no ops person. The first is 50 free, MIT-licensed skills that teach the agent full workflows. The build an outbound prospect list skill, for example, goes from ICP to a send-ready list. The second is integrations that match a small stack: Google Sheets and CSV, Instantly and Smartlead through skills and CSV, list pushes into LinkedCamp campaigns, and HubSpot, Pipedrive and Attio through skills and the API. If you would rather write code, the same tools are a REST API with TypeScript and Python SDKs.
Example: 100 prospects from one prompt
The founder of Parcelry, a five-person startup, connects Leadsets to Claude and asks: "Find 100 heads of RevOps at US B2B SaaS companies with 50 to 200 people that are hiring SDRs. Verified emails only. Put it in a Google Sheet."
- The agent calls
get_usageto check what is left this window. search_companiesfinds US B2B SaaS accounts in the 50 to 200 headcount band.scrape_jobsreads their careers pages and keeps the accounts with open SDR roles. Each standard page read uses one scrape unit.search_peoplefinds revenue leaders at VP or head level at those accounts.enrich_personadds current role and a verified work email. One fictional result: Maya Chen, VP Revenue Operations at Northwind Analytics (northwind.co). The address is maya.chen@northwind.co, with status verified and confidence 0.94. The match reasons are "Revenue operations leader" and "Company hiring SDRs".- The agent writes the verified contacts to one tab, puts catch-alls on a separate tab, and reports what it could not confirm.
A run like this uses roughly one search record per person or company returned and one enrichment record per person enriched. That fits inside the Solo plan's five-hour window. Before you send to the catch-all tab, read what a catch-all email is. Then export the sheet as CSV into your sequencer.
Which plan usually fits
- Free, $0. You get 1,000 enrichment records and 100 search records (10 per call) for life, with no card. Search is capped at 100 records in total, at most 10 per call. Every email is verified. Scraping is not available on Free. This is enough to test one ICP end to end.
- Solo, $29 a month. You get 1,000 enrichments and 500 search records per rolling five hours, and 5,000 enrichments and 2,500 search records per week. Up to 3 searches can run at once, and US personal email is included. This usually fits a founder building a few lists a week.
- Pro, $99 a month. You get 15,000 enrichments and 7,500 search records per five hours, and 75,000 enrichments and 37,500 search records per week. Up to 5 searches can run at once. This suits a small team with a couple of sellers, or a one-off CRM backfill.
- Team ($399) and Max ($999) are for teams that have outgrown the $99 limits and for agents running around the clock.
Scraping is an add-on on any paid plan: $9 for 1,000 units, $19 for 5,000 or $49 for 20,000. Upgrades apply immediately, and you can cancel any time. Full limits are on the pricing page.
What Leadsets does not do yet
- No phone numbers. If cold calling is central to your motion, a provider with phone data is a better fit today.
- SOC 2 is planned, not done. SOC 2 Type I readiness is underway, with a target of Q3 2027. If a customer's security review needs SOC 2 now, choose a vendor that already has it. Today Leadsets runs a GDPR-aligned program, offers a DPA with standard contractual clauses, and has a public opt-out.
- Not a CRM or a sending tool. Leadsets feeds the tools you already use.
- Business data only. There are no consumer contacts, sensitive categories or home addresses.
- No scraping on Free. Hiring signals and site research need a paid plan and scrape units.
You also remain responsible for following the outreach laws that apply to you, such as CAN-SPAM, GDPR and CASL, and for honouring unsubscribes.
Go deeper