Snapchat Public Discover Stories

Extract public Snapchat Discover story, creator, thumbnail, video-track, and bounded snap-media metadata from Snapchat’s anonymously rendered Discover page.

What it returns

Each dataset row is one complete Discover story and can include:

• Story ID, type, title, publication time, tap ID, and result position
• Creator username, display name, public bio/badge, profile URL, profile image, and Snapcode image
• Thumbnail and public video-track URL
• Total snap count plus a bounded snaps array with IDs, media types, media URLs, preview URLs, and timing
• Search context and request provenance: keyword, country, sort order, result page, source URL/status, proxy use, method, and timestamp

Bulky raw __NEXT_DATA__ story objects are not retained. Incomplete stories without a stable ID, title, publication time, thumbnail, and at least one complete snap are excluded.

Input

{
  "keyword": "sports",
  "country": "us",
  "sortBy": "recent",
  "maxItems": 10,
  "maxPages": 1,
  "maxSnaps": 10,
  "requestDelayMs": 250,
  "timeoutMs": 45000,
  "maxResponseBytes": 4000000,
  "proxyConfiguration": {
    "useApifyProxy": false
  }
}

Runs are bounded to 100 stories, one public feed snapshot, and 100 snaps per story. Unknown fields and invalid country/sort values fail before requests begin.

Snapchat’s current public page returns the same server-rendered feed when keyword query values change and does not expose a public cursor for pagination. The Actor therefore requests the canonical /discover page exactly once. keyword, country, and sortBy remain caller-supplied research context, while every record explicitly sets filtersApplied to false and includes a filterNotice.

Access failures and page changes

The Actor opens the public page in Chromium and inspects rendered __NEXT_DATA__ plus JSON responses loaded by that anonymous page. It does not call private or login-only endpoints directly. Access challenges, browser failures, missing state, and empty result sets are recorded in the OUTPUT key’s bounded failures array and counters. They are never mixed into dataset rows.

Chromium uses a stable desktop context, conservative single-page concurrency, bounded retries with Crawlee backoff, response-size/time limits, and direct access by default. Standard Apify Proxy can be enabled when justified. The Actor does not log in, solve CAPTCHAs, patch fingerprints, or bypass access controls.

Responsible use

Use public data only for lawful research, follow Snap terms and robots guidance, and respect creator privacy, copyright, and media rights. Returned URLs do not grant permission to download, republish, or commercially reuse content.

Local validation

Run npm test, apify validate-schema, and apify run --purge --input-file qa-inputs/snapchat-discover-scraper/local-validation.json. Inspect the default dataset and OUTPUT key.