Alibaba RFQ Leads AI: Buying Requests Matched to Your Products

New Alibaba buying requests that fit your products, translated and ready to quote, every morning.

⚡ First 25 buying leads free · 💵 ≈ $0.01 per qualified buying lead all in, source scan included · 🤖 ≈ $0.016 all in with AI, on plans with model access or your key · ⏱️ demo in under 10 s

Try it in 30 seconds. Click Try it — six sample RFQs and an exporter profile are pre-filled. Type your own product lines and press Start; we charge only for the requests that fit. → What a run costs · What the AI adds · Run it daily

⚡ At a glance

What you get	open Alibaba buying requests that fit your catalogue, MOQ and markets — scored, translated, with a quote reply
Typical run	200 requests read → ≈ 67 qualified (1 in 3, estimate until a live run is measured)
Cost of that run, all in	$0.67 rules only · $1.10 with the AI — ≈ $0.01 / $0.016 per qualified buying lead, source scan included
Free tier	your first 25 qualified buying leads on this Actor — one free tier per Apify account, however many searches you run
Keys / setup	the AI runs on Apify's model access (plans that include it) or your own key — a free Gemini key gives 1–3 AI verdicts per run, a paid key every buying lead
Non-English requests	leave the AI on — the rules read English only; the AI translates Spanish, Portuguese, French, Arabic
Which plan	Free demo and your own RFQ rows (list mode, template): any plan · live search: a plan that runs Store Actors
To quote you need	your Alibaba.com supplier account — quotes go out from it, and your membership caps how many a month; quotesPerDay spends that quota on the best leads
Agencies	10 clients daily ≈ 3,000 leads a month ≈ $58 with the AI — vs one go4WorldBusiness Gold+ account at $100 a month

🎯 What this Actor does

• Fit to your catalogue — 0–100 against your products, MOQ, certificates and markets, reason quoted.
• The request read for you — in English (original kept); specs, quantity, target price, incoterm, destination and deadline as columns.
• Ready to answer — hot / warm / cold, a quote today flag, a reply in the buyer's language from your facts, and with Quotes I can send today the requests most worth today's quotes.
• An HS heading suggestion (4–6 digits, labelled a suggestion).
• Only new requests on a schedule — memory by RFQ id; filters and duplicates are never charged by us.

🚀 How to use it

1. Click Try it — the free demo scores six sample RFQs against a sample exporter profile.
2. Type your product lines and company facts — they replace the sample and start the live search, capped by "Max source spend". Leave "AI assessment" on (which AI your plan gets).
3. Press Start — buying leads land best first; "Quote replies" has the text for Alibaba. Quoting needs your Alibaba supplier account: spend your limited quotes on the hot rows first.
4. Schedule it daily with your e-mail in notifyEmail — only new requests are delivered and charged. Working in a spreadsheet? Start from the RFQ inbox sheet (new / quoted / won).

📦 Already scraping Alibaba RFQs? Qualify your dataset

Paste the dataset ID of any Alibaba RFQ scraper into Dataset (sourceMode: dataset) — no second scrape.

🔎 Find buyers on Alibaba without reading every request

It reads the whole RFQ board for your product lines and hands you the few worth a quote — 📚 step-by-step tutorial.

🆚 Why this instead of running the Alibaba RFQ scraper yourself?

	Alibaba RFQ Scraper (blueskyscraper/alibaba-rfq-scraper)	This Actor
Your time per 100 buying leads	≈ 10 h reading 300 rows, translating, writing quotes	≈ 0 h — ranked, read, reply drafted
Requests you never read	none — every keyword match	the ≈ 2 in 3 that do not fit
What you pay for	every row ($0.002), off-catalogue and closed ones too	only buying leads that fit — junk costs $0 here
100 usable buying leads, all in	$0.60 for 300 unsorted rows you read yourself	100 ranked, explained leads: $1.00 rules only · $1.64 with the AI, scan included
What you get	the buyer's language in one HTML-ish string, no answer	0–100 fit with the reason, English + specs as columns, quote reply

vs other Alibaba RFQ scrapers and buy-lead sites

	100 usable buying leads, all in	Your time per 100	Fit + translation + quote draft	Start fee
This Actor	$1.00 rules · $1.64 AI, scan included	≈ 0 h	yes, with the quoted reason	no
blueskyscraper	$0.60 (300 rows × $0.002; 2.7 % of runs failed)	≈ 10 h	no — keyword matches	no
codingfrontend	$1.20 (300 × $0.00399)	≈ 10 h	no; keyword searches may return nothing (its README)	$0.001
apt_marble	$2.40 (300 × $0.008)	≈ 10 h	keyword relevanceScore only (its README)	no
xtracto	$9.00 (300 × $0.03; 7 % of runs failed)	≈ 10 h	quality score only	$0.005
zen-studio Alibaba Scraper	products and suppliers, no buying requests	—	no	$0.005
go4WorldBusiness Silver	$75 a month, 10 leads a day, its own board	hours	no	—
Alibaba RFQ Market	free to browse, one keyword at a time	hours	no	—

💸 A go4WorldBusiness alternative for Alibaba: ≈ 10 new fits a day cost ≈ $3.80 a month on rules or ≈ $5.70 with the AI, source scan and daily digest included — against $75 a month for a Silver seat (read 2026-10-22).

💵 Pricing — what a run really costs

Pay per result: one event per qualified buying lead, plus one digest per run that brought new billed leads. Your first 25 qualified buying leads are free, whatever the tier.

Event	When it is charged	Price
free-tier	your first 25 qualified buying leads on this Actor, in any mode	$0.00
qualified-buying-lead-basic	AI off — normalised request, rule fit score, regex reading, template reply	$0.004
qualified-buying-lead-ai	AI on — plus AI fit with a quoted reason, translation, specs, HS heading, reply in the buyer's language	$0.01
digest-delivered	once per run that billed ≥ 1 new buying lead: the e-mail / chat digest and the report page (first 3 free)	$0.03

Never charged here: requests below minScore, filtered, suppressed or already received. The source's scan is not ours to waive: the RFQ scraper bills your Apify account ≈ $0.002 per request read — capped by maxDiscoveryChargeUsd.

How that compares — $0.004 is the price of one raw row of the Alibaba RFQ feeds (median $0.004), but one qualified buying lead replaces ≈ 3 raw rows: $0.0013 per raw-row equivalent, and the rows you would have dropped cost nothing here.

Worked example — 200 requests read → 67 qualified, AI on: source $0.40 on your account + 67 × $0.01 + 67 × ≈ $0.0004 tokens = $1.10, about $0.016 per qualified buying lead; AI off: $0.40 + 67 × $0.004 = $0.67 ($0.010 each). A run past your free leads adds one $0.03 digest.

At volume — an agency, 10 clients daily at ≈ 10 fits each: ≈ 3,000 leads a month ≈ $58 with the AI, $39 on rules.

AI tier, all in: ≈ $0.0104 per buying lead here ($0.01 + ≈ $0.0004 tokens: 2,051 in / 453 out measured on google:gemini-3.5-flash-lite with an own Gemini key; the default via Apify's model access is not measured yet) — ≈ $0.016 with the source scan.

🤖 What the AI tier adds

The same Spanish request, AI off — rule score, regex reading, an English template reply:

{"subject": "Embroidered hoodies for school uniforms", "language": "es", "descriptionEnglish": null,
 "ruleScore": 92, "label": "hot", "targetPrice": 7.5, "incoterm": "CIF", "destination": "Manzanillo",
 "fitReason": "matches your product line 'hoodie' (hoodie); 1,200 pieces meets your MOQ; Mexico is one of your markets",
 "hsCodeSuggested": null, "replyLanguage": "en"}

AI on (qualified-buying-lead-ai, run of 2026-10-22 on google:gemini-3.5-flash-lite with an own Gemini key):

{"subject": "Embroidered hoodies for school uniforms", "language": "es", "score": 90, "label": "hot",
 "descriptionEnglish": "I need 1200 cotton and polyester hoodies, 320 gsm, sizes S-XL, navy blue, with an embroidered school logo on the chest. Target price 7.50 USD per piece CIF Manzanillo. …",
 "specs": [{"name": "colour", "value": "navy blue", "quote": "color azul marino"}],
 "fitReason": "Matching product: hoodies in cotton-polyester fleece, 320 gsm; Quantity 1200 pcs is above MOQ (500) — \"sudaderas con capucha de algodón y poliéster, 320 gsm\"",
 "hsCodeSuggested": "611020", "hsConfidence": 0.9, "replyLanguage": "es",
 "quoteReply": "Hola, gracias por su solicitud de sudaderas con capucha. En Sunrise Knit Exports fabricamos sudaderas con capucha de algodón y poliéster de 280-340 gsm, …",
 "aiModel": "google:gemini-3.5-flash-lite"}

That is ≈ 6 minutes of reading, translating and drafting per request by hand (on 5 sample buying leads, 2026-10-22). The AI quotes every value; a reply naming a price, certificate or lead time you did not state is dropped.

Which AI your plan gets: Apify's model access where your plan includes it; otherwise your own key (llmProvider: byok) — a free Gemini key (aistudio.google.com/apikey) gives 1–3 AI verdicts per run, a paid key every buying lead. The free demo shows labelled sample verdicts (aiModel: "sample").

⚙️ Input

Two fields start a search — Your product lines (products) and About your company (sellerProfile); minScore decides what you pay.

Every other input — keywords, excludeTerms, draftReplies, sendAlertsTo, flags, sources, webhooks — is in the Input tab and the API reference.

📤 Output

One dataset item per qualified buying lead. Views: Shortlist, Today's quotes, Quote replies, The request read, CRM export (every field flat, AI ones included), Score breakdown.

Sort by score; ruleScore is the rule part, fitScore the AI's — threshold and trend across runs on ruleScore.

Upsert key for a CRM or Sheet: dedupeKey — alibaba-rfq:<rfqId>, the same id whichever term, run or source found the request.

OUTPUT holds the funnel, charges, leadDigest (by country, quantity, specs) and, after a live search, suggestedTerms (words to exclude or add next time); a DIGEST and a client-ready REPORT page sit next to it — with a hot lead, both link a 1688 factory-price check to price it before you quote.

🔁 Run it daily or weekly

1. Open the saved task your first real run links (OUTPUT.scheduleTaskUrl) and click Schedule — daily, as requests fill their quote seats in days. notifyEmail gets the digest: top 10 and the trend vs the last run.
2. Leave dedupeAcrossRuns: true — the next run delivers and charges only what is new.
3. A scheduled run reads 2× maxRequestsToRead, and a spend cap left at $0.50 doubles with it; the status says when a cap stopped the read. Your first live run looks back 30 days.
4. Digest to chat via Send alerts to, rows to your CRM via webhookUrl, a running log via historyDatasetName (tracker sheet).

Daily cost is the new fits only: ≈ 30 new requests read ≈ 10 buying leads — ≈ $0.13 a day on rules, ≈ $0.19 with the AI, scan and digest included.

🎯 Try it for your niche

Pick one in Start from a ready niche search (preset). No preset's yield is measured yet, so each first run is capped at 75 requests and $0.15 of source spend:

Niche	Products it searches	Filters
Garments exporter	cotton t-shirt, hoodie, polo shirt, sweatshirt	1+ buyer stars, 7 days
LED lighting manufacturer	led panel, strip, flood light, bulb	500+ pieces
Spices & food exporter	black pepper, turmeric, cardamom, basmati rice	1+ ton, 14 days
Export agency, 3 clients	cotton t-shirt, led panel light, black pepper	7 days, its own memory

🔌 Integrations, automation and API

Run outcomes — what your integration sees

• Webhook — webhookUrl POSTs {"event": "buying_lead.qualified", "buying_lead": {…}, "runId": "…"} per lead, or {"event": "buying_leads.qualified", "buying_leads": [{…}], "runId": "…"} batches with webhookBatchSize. A 429 / 5xx is retried once after 2 s; a failed POST is retried on the next run ("redelivery": true, up to 7 days, never charged again). webhookSecret signs every POST (X-Signature: sha256=…).
• Which delivery field wins — every URL is posted once per run: Send alerts to fills each service you did not set yourself (an explicit, API-only slackWebhookUrl … wins for its service); webhookUrl gets the leads as JSON, the chat services and notifyEmail the digest.
• HubSpot / Pipedrive — a Make / Zapier step upserted on dedupeKey today; a native crmPush is the next release.
• After every blueskyscraper/alibaba-rfq-scraper run — its Integrations tab: "Run succeeded → start this Actor" with sourceMode: dataset, datasetId = resource.defaultDatasetId, enableAi: false.

MCP & AI agents — add https://mcp.apify.com/?actors=rich_minds/alibaba-rfq-leads-ai to any MCP client. One real item, the OUTPUT keys with types, which call runs what and every flag: API reference.

👥 Who is it for?

You are…	You run it to…	Start with
An exporter / manufacturer (India, Vietnam, Turkey, China)	find buyers for my products each morning, answer first	product lines + profile, daily, quotesPerDay
A sourcing agent / trading company	find importers in my markets, with quantity and price	several product lines, buyerCountries
An export agency	buying leads for each client	the export-agency preset; one saved task per client
A trade platform backend	buyer-intent rows by product and corridor	the API, hsCodeSuggested

🧠 How the AI works

• One typed call per request past the free filters; the model sees the request and your profile only.
• Grounded by code: a value whose quote is not in the request, an HS heading under 0.5 confidence and a reply with facts your profile lacks are dropped; cold rows get no reply.
• Model: google/gemini-2.5-flash-lite via Apify's model access, else your own key; a model failure never fails the run — the rule score stands.

🔒 Data, compliance and limits

• The Store scraper reads Alibaba on your account; this Actor never sends a quote.
• buyerName is a public name as Alibaba shows it; includeBuyerName: false drops it. Mind Alibaba's terms and GDPR.
• Limits: some rows omit description or tags; HS headings are suggestions to check.

❓ FAQ

I already see RFQs in My Alibaba — why this? Your back office shows requests one category at a time and leaves the reading to you. This runs every product line at once, scores each request against your catalogue, MOQ and markets, translates it and drafts the reply — so your limited quotes go to the hot rows first.

Why pay when the Alibaba RFQ Market is free? It wins at browsing the live board, and only there can you send the quote — but one keyword at a time, with no fit, translation, export, schedule or alerts.

Where do the Alibaba buying requests come from? Alibaba's public RFQ board, read by a Store scraper on your account — or a dataset or rows you bring.

How many Alibaba buying leads does a run find? About 1 in 3 requests read (an estimate until a live run is measured): 200 read → ≈ 67.

What does a month of export buying leads cost? A daily schedule at ≈ 10 new fits a day: ≈ $3.80 a month on rules, ≈ $5.70 with the AI, source scan and daily digest included.

Is this a go4WorldBusiness alternative? For Alibaba buying requests, yes — ≈ $5.70 a month with the AI against $75 for a Silver seat.

Where do I find trade leads for exporters? Buyers post them as RFQs on Alibaba's board every day; this reads it for your product lines and keeps the requests that fit your catalogue, MOQ and markets.

Are these real import buyers? Yes — importers, wholesalers and brands who posted a request on Alibaba, with country, quantity and buyer stars; minBuyerStars keeps the serious ones.

Do I need an Alibaba supplier account and quote quota? To answer, yes: quotes go out from your Alibaba.com supplier account, and your membership limits how many you send a month (your RFQ page in My Alibaba shows what is left). Set quotesPerDay to your daily share and the run marks the leads most worth one.

Can an agency run b2b buying leads for several clients? Yes — one saved task with its own dedupeStoreName per client (the export-agency preset shows three); one run for several clients is on the roadmap.

How long does a run take? Demo under 10 s; a free google:gemini-3.5-flash-lite key measured 2026-10-22: ≈ 1.7–13.8 s per buying lead (≈ 4–35 a minute) — 50 leads ≈ 870 s; worst case per 100 ≈ 23 min.

🧩 More Actors from the same developer

• LinkedIn Buying Intent Leads — more buyers for you: people on LinkedIn asking for a product like yours.

🆘 Support

Open an issue on the Actor's page.

⭐ Found it useful? A review on the Store helps other exporters find it.

📝 Changelog

• 0.1.4 (2026-10-23) — If your Apify plan cannot run the live search, the Console now scores the free sample requests against your own products instead of returning nothing; a scheduled run that hits the same wall e-mails / posts the refusal to your channels; Each run that brings new billed buying leads now includes a $0.03 digest (e-mail / chat summary and report page; your first 3 are free), and a hot lead's digest links a 1688 factory-price check to price it before you quote; The page now says up front that quoting needs your Alibaba supplier account and its monthly quote quota, and names the model and key behind every measured AI speed and token figure.
• 0.1.3 (2026-10-23) — sample requests shown in English in the free demo; measured own-key AI speed and the timeout for 50 leads.
• 0.1.2 (2026-10-23) — failed webhook posts re-sent; English subject first; today's quote plan; cheaper AI.
• 0.1.1 (2026-10-23) — initial release.