Pinterest Autocomplete Keyword Scraper

Turn seed phrases and Pinterest search URLs into ranked, export-ready Pinterest autocomplete keywords. The Actor opens Pinterest anonymously, captures its public autocomplete response, removes duplicates, and can recursively expand suggestions into a long-tail keyword map.

Use it to plan Pinterest SEO, seasonal content, ecommerce collections, promoted-pin themes, and client research without copying suggestions by hand. No Pinterest account or private API key is required.

What does Pinterest Autocomplete Keyword Scraper do?

For each input keyword, the Actor:

1. opens the matching public Pinterest pin-search page;
2. captures Pinterest's structured autocomplete response;
3. validates and normalizes the returned suggestions;
4. preserves rank, suggestion ID, type, and destination URL;
5. deduplicates repeated suggestions across inputs;
6. retains all parent seed relationships; and
7. optionally uses discovered suggestions as the next expansion level.

The default depth of 0 performs one fast autocomplete lookup per input. Depth 1 or 2 builds a wider long-tail graph while maxSuggestions keeps the job bounded.

Who is this Pinterest keyword tool for?

• Pinterest marketers discovering phrases people are prompted to search.
• SEO teams comparing Pinterest language with web-search keyword lists.
• Ecommerce planners naming collections, gift guides, and seasonal landing pages.
• Content teams turning broad themes into article, video, and pin-board ideas.
• Agencies refreshing keyword exports for multiple clients or locales.
• Data teams feeding ranked suggestions into Sheets, BigQuery, a warehouse, or an LLM workflow.

Why use autocomplete data?

Autocomplete is a useful source of real topic phrasing because suggestions appear directly in Pinterest's public search experience. It can reveal modifiers such as style, audience, room, season, color, product, and occasion.

This Actor returns source data rather than estimating search volume. It does not claim that a suggestion has a specific monthly volume, trend score, or paid-ad value. Combine it with your analytics and campaign data when prioritizing keywords.

What Pinterest keyword data is extracted?

Field	Meaning
seed	Input or recursive parent that produced the record
parentSeeds	Every processed parent connected to the deduplicated suggestion
suggestion	Normalized autocomplete query
label	Human-readable label returned by Pinterest
suggestionId	Pinterest autocomplete ID, when available
rank	One-based position for this parent seed
resultIndex	Zero-based Pinterest result index
type	Usually query; guide identifies the rendered fallback
searchUrl	Absolute Pinterest search URL for the suggestion
sourceUrl	Search page opened for the parent seed
locale	Requested browser locale
country	Country detected by Pinterest, when available
depth	Parent expansion level; direct inputs are 0
scrapedAt	ISO 8601 collection timestamp

How to scrape Pinterest autocomplete keywords

1. Open the Actor input page.
2. Add up to 100 phrases in Seed keywords, or paste Pinterest pin-search URLs.
3. Choose a browser locale such as en-US or de-DE.
4. Keep expansion depth at 0 for a quick export, or choose 1–2 for long-tail research.
5. Set the global maximum number of unique suggestions.
6. Run the Actor.
7. Export the default dataset as JSON, CSV, Excel, XML, or RSS.

A useful first run is:

{
  "seedKeywords": ["summer wedding", "fall home decor"],
  "locale": "en-US",
  "expansionDepth": 0,
  "maxSuggestions": 50
}

Input parameters

Parameter	Type	Default	Description
seedKeywords	string array	none	Phrases to expand; up to 100 unique values
startUrls	request list	none	Pinterest /search/pins/?q=... URLs
locale	string	en-US	Browser locale and deduplication context
country	string	none	Optional ISO country for a supplied Apify proxy
expansionDepth	integer	0	Recursive levels from 0 through 2
maxSuggestions	integer	100	Global output cap from 1 through 5,000
proxyConfiguration	object	direct	Optional user-controlled Apify Proxy settings

At least one non-empty seed or valid Pinterest search URL is required. Only public Pinterest pin-search URLs are accepted in startUrls.

Pinterest search URL input

If your workflow already stores Pinterest searches, pass them directly:

{
  "startUrls": [
    { "url": "https://www.pinterest.com/search/pins/?q=home%20office%20ideas" }
  ],
  "maxSuggestions": 25
}

The Actor extracts the q value and applies the same locale, limits, deduplication, retry, and expansion behavior used for direct seeds. Unsupported domains and empty q parameters fail clearly instead of producing misleading empty datasets.

Recursive long-tail keyword expansion

Set expansionDepth to 1 to scrape the initial seed and then scrape its discovered suggestions. Depth 2 adds one more level. The queue is breadth-first, so broad coverage is collected before deeper branches.

maxSuggestions limits unique output records, not page visits. A suggestion returned by multiple parents appears once with all relationships in parentSeeds. The first parent remains in seed, and depth shows the level where that record was first saved.

For a bounded research map:

{
  "seedKeywords": ["small balcony garden"],
  "expansionDepth": 2,
  "maxSuggestions": 150
}

Output example

A current anonymous Pinterest response produces records like:

{
  "seed": "summer wedding",
  "parentSeeds": ["summer wedding"],
  "suggestion": "summer wedding guest dress",
  "label": "summer wedding guest dress",
  "suggestionId": "ac_qvTrYIzi",
  "rank": 1,
  "resultIndex": 0,
  "type": "query",
  "searchUrl": "https://www.pinterest.com/search/pins/?q=summer+wedding+guest+dress&rs=ac&len=14&source_id=ac_qvTrYIzi",
  "sourceUrl": "https://www.pinterest.com/search/pins/?q=summer%20wedding",
  "locale": "en-US",
  "country": "DE",
  "depth": 0,
  "scrapedAt": "2026-07-26T07:14:53.029Z"
}

IDs, suggestions, country detection, and rankings come from Pinterest and can change between runs. Fallback guide records can have suggestionId: null.

How much does it cost to scrape Pinterest autocomplete keywords?

The Actor uses pay per event: you pay only for each unique suggestion saved to the dataset. There is no separate start fee. The FREE tier is $0.00075178 per saved suggestion, with lower per-item rates on higher Apify tiers.

Multiply the number of saved suggestions by your plan's displayed per-item rate to estimate a run; for example, 100 results use 100 item events and 1,000 results use 1,000 item events.

Apify rounds and bills according to your platform plan. Optional proxy traffic and platform usage can affect your total account cost. Prices shown here describe this Actor's current source-configured pricing and must be checked on the live pricing tab before a production budget is finalized.

Reliability, retries, and proxy behavior

The implementation listens for Pinterest's autocomplete network response before navigation, validates its success status, and blocks unnecessary images, video, and fonts. Transient failures are retried twice with a fresh isolated browser context and bounded backoff. A visible public search-guide extraction is used only when structured autocomplete does not arrive.

Direct anonymous access is the default. The Actor never enables a residential fallback automatically. If you supply proxyConfiguration, its session and optional country are used for that attempt; proxy costs remain under your control. A terminal blocked or changed response fails the run rather than silently reporting zero keywords.

Limits and responsible scaling

• Maximum 100 unique input seeds per run.
• Expansion depth is capped at 2.
• Output is capped at 5,000 unique suggestions.
• Pinterest commonly returns a small bounded set per keyword.
• Locale and IP geography can influence results.
• Public endpoint shape and availability may change.

For large research jobs, start with depth 0, inspect relevance, and expand only high-value themes. Prefer scheduled incremental refreshes over repeatedly running thousands of low-value variants.

Integrations and automation ideas

• Schedule weekly exports to detect new seasonal phrasing.
• Send dataset items to Google Sheets for editorial triage.
• Join autocomplete phrases with Search Console or paid-search metrics.
• Group modifiers with an LLM after extraction.
• Build Pinterest board and collection naming briefs.
• Trigger Slack or email alerts when a target suggestion appears.
• Export CSV files for client deliverables.
• Use webhooks to launch downstream enrichment Actors.

Run with the Apify API

Replace YOUR_TOKEN with your Apify API token.

cURL

curl -X POST \
  "https://api.apify.com/v2/acts/automation-lab~pinterest-autocomplete-keyword-scraper/runs?token=YOUR_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"seedKeywords":["summer wedding"],"maxSuggestions":25}'

JavaScript

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('automation-lab/pinterest-autocomplete-keyword-scraper').call({
  seedKeywords: ['summer wedding'],
  locale: 'en-US',
  maxSuggestions: 25,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Python

from apify_client import ApifyClient
import os

client = ApifyClient(os.environ['APIFY_TOKEN'])
run = client.actor('automation-lab/pinterest-autocomplete-keyword-scraper').call(run_input={
    'seedKeywords': ['summer wedding'],
    'locale': 'en-US',
    'maxSuggestions': 25,
})
items = client.dataset(run['defaultDatasetId']).list_items().items
print(items)

Use with Apify MCP and AI assistants

Add the Actor to Claude Code:

claude mcp add --transport http apify \
  "https://mcp.apify.com?tools=automation-lab/pinterest-autocomplete-keyword-scraper"

Claude Desktop setup

Add this JSON to Claude Desktop's MCP configuration.

Cursor setup

Use the same HTTP server entry in Cursor's MCP settings.

VS Code setup

Use the same server URL in your VS Code MCP configuration.

{
  "mcpServers": {
    "apify": {
      "url": "https://mcp.apify.com?tools=automation-lab/pinterest-autocomplete-keyword-scraper"
    }
  }
}

Example prompts:

• “Expand minimalist kitchen into 50 ranked Pinterest suggestions.”
• “Create a one-level keyword map for winter wedding and return CSV-ready fields.”
• “Compare parent relationships for small patio ideas and balcony garden.”

Legal and responsible use

This Actor accesses publicly available anonymous Pinterest search suggestions. Use the data only for lawful purposes and follow Pinterest's terms, robots guidance, applicable privacy rules, and your contractual obligations. Do not use the Actor to overload Pinterest, evade access controls, or infer sensitive traits about individuals. You are responsible for deciding whether your collection and downstream use are permitted in your jurisdiction.

Troubleshooting

Why did my run fail instead of returning an empty dataset?

A blocked page, malformed response, invalid URL, or exhausted retry path is treated as an error. This prevents upstream changes from looking like a legitimate “no suggestions” result. Check the run log, reduce input size, confirm the URL shape, or supply an appropriate proxy configuration.

Why are suggestions different from my browser?

Pinterest can vary results by time, locale, IP country, experiments, and session context. Match the locale and, when necessary, use a proxy configured for the intended country. The output country field shows what Pinterest reported when available.

Why are there fewer rows than maxSuggestions?

The value is a ceiling, not a promise. Pinterest may return fewer unique suggestions, especially at depth 0, and duplicate terms are merged across parents. Increase expansion depth or add materially different seeds for broader coverage.

Why is a suggestion ID null?

The structured response normally supplies an ID. If the Actor used the visible search-guide fallback, the public link may not expose one; those records use type: "guide" and suggestionId: null.

FAQ

Does this scrape pins or boards?

No. It exports autocomplete keyword suggestions and their search links, not pin or board records.

Is a Pinterest login required?

No. The current workflow uses the public anonymous search experience.

Does the Actor estimate search volume?

No. It preserves Pinterest's ordering but does not invent volume, CPC, competition, or trend metrics.

Can I use multiple locales?

Run separate tasks per locale so each dataset has a clear locale context and predictable deduplication.

Can I schedule it?

Yes. Use Apify Schedules and export or webhook integrations to refresh keyword research automatically.

Related automation-lab Actors

• Pinterest Scraper — collect public pin and profile data after keyword discovery.
• Pinterest Trends Scraper — monitor Pinterest trend surfaces for adjacent research.

Use this Actor first for keyword expansion, then pass selected searchUrl values into a pin-focused workflow when you need content-level data.