AgentData vs Lusha: company intelligence with full tech stacks, no credits
Lusha for direct dials and a polished extension; AgentData for the company, its stack and its people in one free lookup.
Lusha is a contact-first tool with a strong Chrome extension, 290M+ contacts, direct dials and one of the cleaner compliance records in the category. It runs on credits: 40 free a month, then Starter from $37.45 a seat on annual billing ($49.90 monthly) for 400 credits, with an email costing one credit and a phone five; API calls cost at least one credit even when nothing is returned. AgentData starts from the company instead: profile, 3,968-technology tech stack, verified emails, people and switch signals in one lookup, 1,000 a day, $0, API and MCP included. Both have MCP servers. Verified 17 September 2026.
Lusha does the thing it does very well: you're on a LinkedIn profile, you click the extension, you get a number. If your day is phone calls, that's worth paying for and we can't replace it. Our users tend to start one step earlier: which companies should I even be looking at? Lusha's answer to that is a prospecting filter on firmographics; ours is the tech stack, the signals and the company's own website, free, with the people attached. I'd rather you used Lusha for the dial and AgentData for the decision than pretend one tool does both.
The verdict
Choose AgentData if you…- build lists by what companies run, not just who works there
- want the company profile, tech stack, verified emails and people in one lookup without a credit per field
- are building an agent and want an MCP server and API with no minimum-credit-per-request rule
- work in the UK or EU and want profiles built from the company's own site
- want public company and technology pages your team can browse
- need direct dials and mobile numbers with a published accuracy claim (86%)
- prospect inside LinkedIn all day and want the Chrome extension
- want 26 buying signals, lookalike prospecting and job-change alerts in one contact database
- need ISO 27701 and SOC 2 Type II paperwork for procurement
One company, two responses
Lusha's Company API is two calls: search returns what fields exist and how many credits a reveal costs, enrich returns the record for one credit. AgentData's lookup returns the company, technologies, verified emails and people in one call for one lookup. Both structures are from the vendors' own documentation, read 17 September 2026.
{
"data": [{
"id": "v3.eyJjb21wYW55SWQiOiAzMzIyMjY3OH0=",
"name": "Lusha", "domain": "lusha.com",
"size": "51-200", "revenue": "$5M - $10M", "yearFounded": 2016,
"industry": "Business Services",
"locations": [{ "type": "HQ", "country": "United States", "state": "California", "city": "San Francisco" }],
"technologies": ["Salesforce", "HubSpot"],
"openJobs": 12
}],
"billing": { "creditsCharged": 1, "action": "reveal_company" }
}1 credit per company revealed, after a free search call. Note the technologies array: Lusha does return some technographics, as flat names without category or confidence. “A minimum of 1 credit applies per request even if no data is returned.” Up to 100 companies per request; 25 requests/second per endpoint.
{
"domain": "stripe.com",
"company": {
"name": "Stripe", "sector": "Financial Services & Fintech",
"vertical": "Payments", "b2b_b2c": "b2b", "size": "large",
"headquarters_country": "US", "email_pattern": "first.last",
"signals": { "has_pricing_page": true, "has_api_docs": true, "has_careers_page": true }
},
"technologies": [
{ "name": "React", "category": "Framework", "confidence": "high" },
{ "name": "Segment", "category": "Analytics", "confidence": "high" }
],
"emails": { "total": 42, "verified": 38,
"data": [{ "email": "jane.smith@stripe.com", "title": "VP Engineering",
"confidence": 95, "verification_status": "valid", "verified_at": "2026-09-01T10:30:00Z" }] },
"people": { "total": 156, "data": [{ "name": "…", "title": "CTO", "seniority": "c_level" }] },
"lookup": { "consumed": true, "remaining": 999 }
}One lookup: profile, technologies with category and confidence, verified emails, people, web signals. No credits.
Lusha returns revenue range and open job count, which we don't. We return web signals, switch signals and verification status per email, which Lusha's company object doesn't.
AgentData vs Lusha at a glance
AgentDataVerified 17 September 2026 against docs.lusha.com, lusha.com/pricing FAQ, lusha.com/data and lusha.com/mcp-server-ai-data-integration. Plan prices are from Costbench (29 Jul 2026) and Zeliq (3 Aug 2026) because Lusha's pricing table is client-rendered and slider-dependent; confirm on lusha.com/pricing.
Why agent builders and growth teams pick AgentData over Lusha
Lusha's unit is a contact reveal. Ours is the whole company, with its people attached.
Profile, 3,968 technologies, verified emails and people in one lookupLusha returns a flat list of technology names. We return what, in which category, how sure, and when it changed.
technologies[] with category and confidence; switch signals at /signalsBoth have MCP. Lusha's charges a credit even when a request returns nothing.
Hosted MCP + npm, seven tools, freeclaude mcp add agentdata --transport http https://mcp.agentdata.run/mcp \
--header "Authorization: Bearer YOUR_API_KEY"
# then: "Find UK companies running Salesforce and Gong with 200–1,000 staff; give me the VP Sales at each with a verified email."lookup_companysearch_companiesfind_peopleget_technologiesget_signalscheck_usageForty credits is forty emails. One lookup here is every verified email we hold for the company.
1,000 lookups a day, all fields, no seatscurl -H "Authorization: Bearer $KEY" \
"https://agentdata.run/api/v1/lookup?domain=stripe.com"Lusha's company pages return 410 Gone. Ours are the product.
/company/[domain], /tech/[slug], /companies/[sector]Pricing: AgentData vs Lusha
Lusha is priced per seat by credits, with a slider that changes the price. The calculator shows the plan a single seat needs for a monthly credit volume.
You keep at least $449.4/yr with AgentData at this volume.
Lusha prices from Costbench (29 Jul 2026) and Zeliq (3 Aug 2026) teardowns; Lusha's pricing page is client-rendered and slider-dependent, and the Jul/Aug teardowns disagree on Pro annual ($49.90 vs $45.45). Credit costs from lusha.com/pricing FAQ, 17 September 2026: email 1, phone 5. AgentData: 1,000 lookups a day.
Lusha pricing in 2026
- January 2026 (UpLead, secondary): Pro $22.45/user/mo annual with 3,000 credits/user/yr and 3 seats; Premium $52.45 annual, 7,200 credits/yr, 5 seats.
- April 2026 (Cognism blog, secondary): Free 40 credits; Pro $29.90/mo for 250 credits; Premium $69.90 for 600; Scale custom.
- July–September 2026 (Costbench, Zeliq, live FAQ): new Starter tier at $49.90 monthly / $37.45 annual for 400 credits; Pro $69.90 for 600 (the old Premium price point); Premium $399.90 monthly / from $259.95 annual with credits on a slider. Entry paid price up from $29.90 to $49.90 monthly, a 67% rise, exact date unverified.
- No archive captures of lusha.com/pricing exist for 2021–2025 on Arquivo.pt, so no multi-year table is shown.
- AgentData: $19 / $49 / $99 specced April 2026, never launched; free since May 2026.
AgentData vs Lusha, feature by feature
Contacts and dials: Lusha's strongest card
290M contacts with mobiles and a published accuracy claim is a different job from our crawl.
- AgentData: SMTP-verified emails with confidence and verified_at; people with seniority; no phones.
- Lusha: 290M+ contacts, 165M+ emails, direct dials; 98% email and 86% phone accuracy (vendor claims); weekly refresh.
Source: lusha.com/data, 17 September 2026
Technographics
Both return technologies; only one tells you the category, the confidence, and the change.
- AgentData: 3,968 technologies with category and confidence; switch signals per company; browse by technology at /tech.
- Lusha: technologies[] as flat names in the company enrich response.
Source: docs.lusha.com/apis/openapi/company-search-and-enrich, 17 September 2026
Credits and the one-credit minimum
The rule that matters for agents.
- AgentData: No credits. 1,000 lookups a day; a lookup that finds nothing still counts as a lookup, but there is nothing to buy.
- Lusha: Email 1, phone 5, company 1; “A minimum of 1 credit applies per request even if no data is returned”; monthly rollover to 2×, annual reset.
Source: docs.lusha.com user guide; lusha.com/pricing FAQ, 17 September 2026
Agent access
Both ship MCP. Ours is free of credits and open to any account.
- AgentData: mcp.agentdata.run/mcp or npm; seven tools; free.
- Lusha: mcp.lusha.com, 24 tools, OAuth (Claude, ChatGPT, Codex) or x-api-key (Cursor, VS Code, n8n); uses plan credits; Lusha's own pages say both “any plan, including Free” and “Sales Starter plans and above”.
claude mcp add agentdata --transport http https://mcp.agentdata.run/mcp \
--header "Authorization: Bearer YOUR_API_KEY"Source: docs.lusha.com/mcp-docs; lusha.com/mcp-server-ai-data-integration; lusha.com/blog (webinar recap, updated 21 Jul 2026)
Who moves from Lusha to AgentData
Your ICP is “mid-market companies on Salesforce and Gong”. Lusha can filter firmographics and reveal contacts one credit at a time. AgentData finds the companies by what they run and returns the people with them.
An agent that probes a hundred domains and finds forty empty costs a hundred Lusha credits. Here it costs nothing.
Lusha's 42.8M European contacts are real, but its profiles are aggregated. Ours come from the company's own website, so a 30-person Manchester agency is profiled the same way as a San Francisco unicorn.
Switching from Lusha
- Export from Lusha (Lists → Export). Keep the company domain; that's the join key.
- Re-enrich by domain in AgentData via /api/v1/lookup or the MCP; compare verified emails and confidence against Lusha's on 100 companies you know.
- Keep Lusha for direct dials if you call. Monthly plans cancel any time; annual credits reset at year end so time the switch to the renewal.
Field map: Lusha → AgentData
Where Lusha fits better
- You call. Direct dials with a published 86% accuracy claim, five credits each, is the core of Lusha and we don't do phones.
- You prospect inside LinkedIn. The Chrome extension is the workflow; we're API and web app.
- You want buying signals on contacts: 26 signals, job-change alerts, lookalike prospecting, refreshed weekly.
- Procurement needs ISO 27701 and SOC 2 Type II certificates on file.
We would rather you pick the right tool than ours for the wrong reasons.
Frequently asked questions
Related reading
Switching from Lusha takes an afternoon.
128M companies. 3,968 technologies. 12M email addresses. Free.