Twitter Lead Generation: AI Buying-Intent X Leads, 25 Free

People on X asking for a tool like yours — with a reply draft, every week.

⚡ First 25 leads free · 💵 $0.0025 per qualified lead · 🤖 $0.009 with AI assessment · ⏱️ demo in seconds

Try it in 30 seconds. Click Try it — the form is already filled with a working example. Your first 25 qualified leads are free, and tweets that fail your filters are never charged. → What a run costs · What the AI adds · Run it weekly

⚡ At a glance

What you get	people on X asking for a product like yours: their need quoted, role and company, a lead score and a reply draft
You provide	1–10 words buyers use for what you sell (email marketing) and one sentence about your offer
Output	JSON / CSV / Excel dataset, best-first, plus the funnel in the OUTPUT record
Typical run	500 tweets scanned → ≈ 20 qualified leads in ≈ 2 min (estimate; every run measures OUTPUT.timing)
Cost of that run	$0.13 rules only · $0.36 with the AI verdict ($0.075 X Tweet Scraper + 20 × $0.0025 / 20 × $0.009 here + ≈ $0.10 AI tokens), after your 25 free leads
Free tier	first 25 qualified leads per account
Keys / setup	no API key where your Apify plan includes model access — otherwise your own Gemini / Groq key (llmProvider: byok); the free demo shows sample AI verdicts on any plan
Works with	Schedules, webhooks, Google Sheets, Slack, Discord, Make / n8n / Zapier, MCP & AI agents

🎯 What this Actor does

Twitter lead generation and X lead generation for founders, SDRs and agencies — you pay only for people who want what you sell:

• Searches X for buyers, not mentions — "looking for", "any recommendations", "alternative to", "sick of" … queries.
• Drops the junk for free — vendor promos, bots, hiring posts, giveaways, hashtag spam and retweets, each with a reason.
• Tells you who is asking — role, company, website, public e-mail from the bio, fit vs icpBioKeywords.
• Hands you the reply — a ≤ 280-character draft, a one-click replyUrl, isUnanswered to answer first.
• Only new people on every run — memory per tweet and per author.

🧲 Find customers on Twitter who already ask

People tweeting "can anyone recommend…" or "switching from…" are ready to buy — this finds, ranks and answers them.

🔎 Twitter sales prospecting: the queries are built for you

productKeywords (or your offer alone) become X queries without retweets, promo links or your own accounts; competitorNames (Mailchimp, @klaviyo) adds "alternative / switching from / cancel" searches.

🎚️ Twitter buying intent, filtered your way

intentTypes, follower bands, minScore, targetFlags and suppressionList — free, before any charge. Reply drafts name your product only where someone asked for a recommendation or an alternative.

🚀 How to use it

1. Click Try it — the free demo on 9 sample tweets: 4 leads, 5 junk tweets rejected, $0.
2. Type your own offer (plus productKeywords if you know the words — else they are read from the offer and listed in the status) — that switches to the live X search on your account. Leave enableAi on.
3. Press Start — the Reply queue tab lists who to answer, the draft and the reply link.
4. Schedule it with your address in notifyEmail — only new leads are delivered and charged.

📚 Learn more: Step-by-step tutorial — twitter lead generation with this Actor

🆚 Why this instead of an X Tweet Scraper?

	X Tweet Scraper (xquik/x-tweet-scraper)	This Actor
Price	$0.00015 per tweet · 1,386 users / 30 days	$0.0025 per qualified lead, first 25 free
What you pay for	every tweet — promos, bots, jokes included	only qualified leads — junk costs $0
Same 100 leads, all in	$0.38 for ≈ 2,500 tweets, then you read all 2,500	100 ranked, explained leads: $0.63 rules only · $1.78 with AI ($0.25 / $0.90 here + $0.38 source you pay either way + ≈ $0.50 tokens)
Who is buying?	raw rows	intent, stage, quoted need, role, ICP fit
Next step	none	reply draft + one-click reply link

Store neighbours for "twitter lead generation" — ours first; 100 clean leads ≈ 2,500 raw tweets:

Actor	Users / 30 d	Price	100 clean leads	What you still do
This Actor	new	$0.0025 / lead, $0.009 with AI	$0.63 · $1.78 with AI	reply
b2b_leads/x-twitter-lead-intent-finder-scored-buying-intent-leads	1	$0.0005 / scored lead + start	≈ $0.05 + start (100 of its own leads)	phrase-match scoring, no AI verdict or reply draft; Free-plan runs end empty (its README)
xquik/x-twitter-tweet-classifier	1	$0.0003 / tweet	≈ $0.75	write prompts; 11.1 % fail
xquik/x-twitter-brand-monitoring	1	$0.0003 / tweet	≈ $0.75	no lead verdict
api-ninja/x-twitter-advanced-search	429	$0.00035–0.015 / tweet + start	$0.88–37.50	read every row
apidojo/tweet-scraper	8,011	$0.0004 / tweet	≈ $1.00	read every row
kaitoeasyapi/twitter-x-data-tweet-scraper-pay-per-result-cheapest	4,464	$0.00025 / tweet	≈ $0.63	read every row
apidojo/twitter-scraper-lite	2,098	$0.0004+ / tweet + query fees	≈ $1.00+	read every row

The same search also lists follower / profile scrapers (x_guru/twitter-follower-scraper): follower lists, no intent. b2b_leads is cheaper per lead; we cost more because each lead carries an AI verdict (on our 9-tweet demo sample the free rules alone drop the promo, the hiring post, the giveaway and the bot), the quoted need and a reply draft.

Free and SaaS alternatives: X Advanced Search is free but has no scoring, junk removal, export or schedule. Octolens and Brand24 are monthly seats (pricing, prices) — here you pay per new lead only.

💵 Pricing — what a run really costs

Pay per result: one event per qualified lead, nothing else.

Event	When it is charged	Price
free-tier	your first 25 qualified leads on this Actor, in any mode	$0.00
qualified-lead-basic	AI off — intent score, bio parser, lead score, template reply	$0.0025
qualified-lead-ai	AI on — plus the AI verdict, quoted need, ICP fit and a tailored reply	$0.009

One free tier: your first 25 qualified leads are free, whatever the tier.

Never charged: the search, retweets, bots, vendor promos, hiring posts, giveaways, off-topic tweets, leads below minScore, tweets the AI calls "not a genuine buyer", leads you already received. Compute is included.

How that compares — $0.0025 is 1.4× the $0.001745 median of the same-job X Actors the Store shows for "twitter lead generation" (one per developer), and 8× one raw tweet ($0.0003 median) — but one qualified lead replaces 25 raw tweets you would read and judge: $0.0001 per raw-row equivalent.

Worked example — 500 tweets scanned → 20 qualified leads with AI on: X Tweet Scraper $0.075 on your account + 20 × $0.009 here + ≈ $0.10 tokens → $0.36 total, ≈ $0.018 all-in per lead. Rules only: $0.075 + 20 × $0.0025 = $0.13, ≈ $0.006 all-in per lead. Your first 25 leads cost $0 here.

AI tier, all in: ≈ $0.014 per lead ($0.009 here + ≈ $0.005 tokens through Apify's model access or your own key) — 3.6× the basic price before tokens.

🤖 What the AI tier adds

The same demo tweet, AI off — the free rules (intent lexicon, bio parser, template reply):

{"authorUsername": "kaylaships", "intentType": "alternative_seeking", "need": "Seriously looking for a Klaviyo alternative for email marketing before our renewal in October.", "authorRole": "Head of Growth", "intentScore": 83, "icpFit": 83, "leadScore": 82, "leadTier": "hot", "suggestedReply": "@kaylaships Email marketing tool for Shopify stores — happy to share how it handles \"looking for a Klaviyo alternative for email marketing before our renewal in October\"."}

AI on (qualified-lead-ai, $0.009) — the demo's labelled sample verdict for the same tweet (what Try it shows on a plan without model access; your own search gets a live verdict from the default model):

{
  "authorUsername": "kaylaships", "intentType": "alternative_seeking", "buyingStage": "evaluating",
  "need": "looking for a Klaviyo alternative for email marketing before our renewal in October",
  "intentScore": 94, "icpFit": 86, "leadScore": 89, "leadTier": "hot",
  "reasons": ["renewal in October = a deadline", "\"flows keep breaking\" is the pain", "Head of Growth, ecommerce bio"],
  "suggestedReply": "@kaylaships Flows breaking right before a renewal is the worst timing. We built an email marketing tool for Shopify stores with simpler flows — happy to show how a Klaviyo switch works before October.",
  "aiAssessment": {"intentType": "alternative_seeking", "isGenuineBuyer": true, "stage": "evaluating", "confidence": 0.92, "icpFit": 86}
}

The AI tells a real shopper from a vendor or a joke and writes the reply — the 20 minutes a day of reading you skip.

Try the AI on any plan: the free demo shows labelled sample AI verdicts (aiModel: "sample"). Without model access, your first real leads can get live verdicts on the developer's key (OUTPUT.aiTrial). Then your own key (llmProvider: byok): a free Gemini key (aistudio.google.com/apikey) assesses a few leads per run; a paid Gemini / Groq key (billing on, ≈ $0.005 per lead) assesses every lead.

⚙️ Input

{"productKeywords": ["invoicing app"], "offer": "Invoicing app for freelance designers", "competitorNames": ["FreshBooks"], "sourceMode": "actor", "maxDiscoveryChargeUsd": 0.5}

Input options

Every other option is explained in the Input tab.

Flags

Put the name from the first column into targetFlags (decisionMaker works too); an unknown name stops the run with the list of valid ones.

📤 Output

One dataset item per qualified lead — JSON, CSV or Excel, or a webhook.

Output fields

Upsert key for a CRM or Sheet: dedupeKey — tweet:<tweetId>, the key the cross-run memory uses, so one tweet is one row however many runs find it; authorId groups one person's tweets.

OUTPUT = the full run record; RUN_SUMMARY = the same record, same keys. Next to it: a DIGEST (Markdown for Slack / e-mail: counts per intent type, "vs last run: qualified 41 → 55, hot 5 → 9", top leads with links) and a client-ready REPORT page (OUTPUT.reportUrl). Views: Shortlist, Reply queue, Authors (CRM sheet), Intent analysis. OUTPUT holds queries, tweetsScanned, rejectedByReason, leadsByTier, leadsByIntent, sourceCostEstimateUsd, aiCoverage, termsFromOffer, scheduledSizing, sourceError, aiError.

⭐ Found it useful? A review on the Store helps others find it — it takes a minute on the Actor's page.

🔁 Run it weekly

1. Open the saved task and click Schedule — your first real run saves its search as a task and links it in the status (OUTPUT.scheduleTaskUrl). Add notifyEmail: every run with new leads e-mails the digest and the trend.
2. Keep dedupeAcrossRuns: true — tweets and authors you were charged for are remembered per search, not per offer wording. Several clients: one dedupeStoreName each.
3. Add slackWebhookUrl, discordWebhookUrl or webhookUrl / Google Sheets for the rows themselves.
4. From run two, OUTPUT.history and the REPORT page show every run (qualified, hot, unanswered, switchers).
5. A scheduled run covers the days since the last run and scans 1,000 tweets, capped by maxDiscoveryChargeUsd.

Weekly cost is the new leads only, at the page's one hit rate: 1,000 tweets a week → ≈ 40 new leads in week 1 → 40 × $0.009 = $0.36 here + $0.15 of X Tweet Scraper on your account. From week 2 fewer are new: the memory drops the tweets and authors week 1 delivered (the measured week-2 figure follows the first scheduled runs).

🎯 Try it for your niche

Copy the input into the form (JSON view), run it once, then schedule it.

Niche	What it looks for	Start
SaaS founder (invoicing app)	invoicing-app requests, FreshBooks / HoneyBook switchers	input below
B2B SDR (CRM)	CRM requests from sales leaders, score ≥ 70	input below
Competitor conquesting (PM tool)	Asana / Monday.com / ClickUp complaints and switchers	input below

🔌 Integrations, automation and API

• Webhook — webhookUrl (+ webhookHeaders) POSTs to Zapier, Make, n8n or a CRM: {"event": "lead.qualified", "lead": {…}, "runId": "…"} per lead (webhookBatchSize: 1), else {"event": "leads.qualified", "leads": [{…}, …]} per batch. A 429 / 5xx is retried once; a failed POST is counted in OUTPUT.webhook.failed, never fails the run.
• Google Sheets / Slack — one click from the Integrations tab; the Authors view maps onto a sheet.
• HubSpot — Make / Zapier "create or update contact" on the webhook: contact = authorName + authorWebsite + authorRole, note = text + need + suggestedReply; upsert on dedupeKey.
• Pipedrive — the same flow as a person + deal: person = authorName / authorCompany, note = need + suggestedReply, url as the source link; upsert on dedupeKey.
• AI agents / MCP — every input is a plain typed field. First call = the free trial: "sourceMode": "list" with the tweets you send, no search on your account (the snippets below); {} runs the same free demo.

Use it from Claude, ChatGPT or any MCP client — add Apify's MCP server with this Actor as a tool:

{"mcpServers": {"apify": {"url": "https://mcp.apify.com/?actors=rich_minds/twitter-lead-generation-ai"}}}

Then ask: "Run twitter-lead-generation-ai in list mode on these tweets first (free), show me the hot leads, then search with sourceMode: actor, my productKeywords and maxDiscoveryChargeUsd: 0.5."

List-mode rows — what a pasted record needs

sourceMode: "list" (and dataset) run on any Apify plan: rows in the xquik/x-tweet-scraper flat shape, or apidojo/tweet-scraper / kaitoeasyapi output with its nested author object.

Run outcomes — what your integration sees

👥 Who is it for?

You are…	You run it to…	Start with
A SaaS founder / indie hacker	find people asking for a tool like yours today and answer first	productKeywords + offer, AI on
An SDR or growth marketer (B2B)	a daily ranked list of in-market prospects with role and company, in Slack or the CRM	+ icpBioKeywords, minScore: 70
An agency	one saved task per client, scheduled, webhook to each client's pipeline	one dedupeStoreName + webhookUrl per client
A competitor-conquesting team	people complaining about or leaving a named competitor	competitorNames, intentTypes: ["alternative_seeking", "pain_point"]

🧠 How the AI works

• Grounded. One tweet + the public profile per call; the need must be a verbatim quote and role / company must appear in the bio, else the rule value or null is kept.
• Replies are checked. A link other than offerUrl, a number not in the tweet or offer, or a "happy customer" claim → the template; weak leads and non-buyers get none.
• Tokens only where they matter. Every tweet past the free filters (rule intent ≥ 30), best rule score first, capped at aiCandidateMultiplier × maxQualified.
• Model. anthropic/claude-haiku-4.5 via Apify's model access, or llmProvider: "byok" with your own key; if it fails, leads keep the rule verdict.

🔒 Data, compliance and limits

• Public tweets and profiles only, via X Tweet Scraper on your account — no X login; an e-mail only when the author published it in the bio. Not affiliated with X Corp. The Actor never posts — replies are drafts you send.
• X search returns what X indexes; niche terms can return few tweets. Use the output within X's terms and GDPR / CCPA.

❓ FAQ

How much will one run cost me? 500 tweets ≈ $0.075 on your account, capped by maxDiscoveryChargeUsd ($0.50); ≈ 20 leads × $0.009 with AI (× $0.0025 without) here, the first 25 free, plus ≈ $0.005 tokens each.

Is this an X tweet scraper? It runs xquik/x-tweet-scraper (X Tweet Scraper) for you and keeps only tweets from people who want what you sell — you never pay us for the rest.

Where do the twitter leads come from? Public tweets asking for a tool like yours — one person, one tweet, the need quoted; never a bought list.

Why pay when X Advanced Search is free? For a one-off look at a live thread it wins. It has no scoring, junk removal, export, API, schedule or memory — you read every tweet, every day. This is the X Advanced Search alternative for doing it daily.

How is this an Octolens alternative? Octolens and Brand24 sell monthly seats; here you pay per new qualified lead, with a reply draft, and nothing when nothing new shows up. Octolens also watches Reddit, LinkedIn and Hacker News: this Actor covers X; LinkedIn is LinkedIn Buying Intent Leads; Reddit and Hacker News are on the roadmap.

Which Apify plan do I need? Any plan runs the free demo. The live X search and Apify's model access need a plan that can run Store Actors; on the free plan use dataset / list mode and llmProvider: "byok" — a free key assesses a few leads per run, a paid Gemini / Groq key (billing on) every lead.

🧩 More Actors from the same developer

Not quite your use case? The same pay-per-qualified-result model, closest to this Actor's buyers first:

• LinkedIn Buying Intent Leads — linkedin buying intent, linkedin intent signals, abm intent
• Link Building Prospects & Guest Post Finder (Google, AI) — backlink prospecting, link prospecting tool, link building prospects

🆘 Support

Something missing or wrong? Open an issue on the Actor's page — buyer requests ship first.

📝 Changelog

One line per published build, newest first — what changed for you: output, status messages, prices.

• 0.1.4 (2026-09-27) — 2026-09-27.md — M-01, M-02.
• 0.1.3 (2026-09-27) — your offer alone starts a live search (terms read from it, OUTPUT.termsFromOffer); scheduled runs cover the days since the last run and scan 1,000 tweets (OUTPUT.scheduledSizing); views and snippets use score / label; the status says only "N free leads left".
• 0.1.2 (2026-09-27) — AI on without model access: your first real leads get live verdicts; a live run saves its search as a task you schedule in one click; the digest and report name the LinkedIn sibling; HubSpot / Pipedrive recipe.
• 0.1.1 (2026-09-27) — initial release: buying-intent leads from X with free junk filters, the quoted need, ICP fit and a reply draft.