archive.org scrapers compared

Single result in the default dataset.

• Additional cost for monitoring mode (additional-data)
• Agent direct phone delivered (contact-phone)
• AI market analysis (market-analysis)

Per dataset item returned

One ad record (one promoted tweet) pushed to the dataset — with spend, impressions, targeting, billing and/or hydrated tweet creative where available.

• Additional data (additional-data)

One resolved source-post manifest saved, including a truthful partial manifest; unavailable/invalid/not-attempted inputs are not charged

• 10 MiB stored (stored-10-mib)

Single result in the default dataset.

• Content downloaded (content-downloaded)

One clip or short VOD, up to 200 MB.

• Video (medium) (video-medium)
• Video (large) (video-large)
• Video (XL) (video-xl)
• Video (2XL) (video-2xl)

One review row in the dataset.

Single result in the default dataset.

• Comment thread (comment-thread)

Charged per item extracted.

• Run started (start)

One snapshot, latest-snapshot, or search result row

Charged for each chat message written to the dataset.

• Actor start (actor-start)

Each URL checked against the Wayback Machine

One story item from a profile's highlights, with media URLs, link stickers, mentions, hashtags, locations, music, sponsor tags and dates.

• Media archived (media-archived)
• Profile checked, nothing new (empty-check)

One public archived Reddit post or comment is saved to your dataset.

One post or comment returned

One row pushed to the dataset — a Wayback snapshot, a library search result, an item-metadata record, or an availability check. $1.50 per 1,000 rows.

• Additional data (additional-data)

Cost per record scraped.

One Reddit Best of archive post saved to the default dataset.

One validated historical Reddit post or comment persisted with RAG text, canonical source URL, creation time, and archive provenance.

Charged per chat message pushed to the dataset.

• Actor start (actor-start)

One unique original PDF saved in the run ZIP. Failed inputs and duplicates have no item charge.

• Run started (start)

One Substack post or one Substack Note written to the dataset.

One saved unique dataset record.

One newsletter post row written to the dataset. Paid posts include public previews only. Error rows are not charged.

1 Substack post's metadata collected ($1.00 per 1,000)

One bounded public-source request attempt; charged also when access fails or filters match nothing.

• Run initialization (including invalid inputs) (run-start)
• Delivered post record (result)

One public post written to the dataset with its metadata, body and engagement counts. Error rows are not charged.

• Scraped comment (comment-scraped)

Charged per post emitted (metadata + full text). Paywalled posts return the public preview.

• Comment scraped (comment)
• Publication profile (publication)

Cost per record scraped.

• Actor start (orbtop-actor-start)

One URL checked in the Wayback Machine (first, latest and closest snapshot with links)

One successful file metadata observation row written to the default dataset.

One unique archive_item metadata row saved to the dataset. Repeated IDs, source failures and the archive_search_summary are not charged.

Charged per archive.org item returned, from a search or an item lookup.

One article with full text, byline, tags, and metadata.

• Full article text (article-full-text)

Cost per record scraped.

Single result in the default dataset.

• some event (some-event)

One successful selected metadata observation saved to the dataset.

One archive.org item with metadata and links.

• File (file)

One archive item delivered. Searches without results and rejected queries are never charged.

One scraped record. No start fee, no minimum; failed or empty runs charge nothing.