Instagram Reel Details

Collect details for public Instagram Reels, including captions, creator information, and video/audio metadata. The output contains normalized fields and a sanitized nested media object; field availability varies by Reel. This is not the full HTTP response or unrelated feed data.

Published on the Apify Store. Cloud collection has been tested. Pay-per-event billing is active for successful default-dataset items; failed observations go to the run summary and do not trigger the result event.

Input

{"urls": ["https://www.instagram.com/reel/SHORTCODE/"]}

Replace SHORTCODE with a real public Reel identifier. Supply 1–1,000 URLs; duplicates and tracking-query variants are collected once. /reel/SHORTCODE/ and /USERNAME/reel/SHORTCODE/ URLs are supported; any other URL is skipped, listed under invalidUrls in SUMMARY, and not billed. No customer Instagram credentials are required.

Output

Field	Meaning / source
schemaVersion	2; full-media response contract
url, shortcode, mediaId	Canonical Reel URL and identifiers; IDs are strings
collectedAt	Actual capture timestamp, ISO 8601 UTC
playCount	Instagram play_count; not unique viewers
instagramPlayCount	Separate ig_play_count; never substituted for play_count
likeCount, commentCount	Reported likes and comment counts
reshareCount, saveCount	Reported reshare_count and save_count; conditional availability
caption, hashtags	Caption text and hashtags extracted from it
takenAt	Source publication timestamp in Unix seconds
author	ID, username, display name, verified flag and profile picture URL
video	Duration in seconds, original width/height, video versions, image candidates, audio flag
location, coauthors	Source location and coauthor data when present
audio	Music info or original-sound info from clips metadata
raw	Captured media object, including additional fields and nested structures
removedFields	Paths removed from raw; values are never included

Missing normalized scalar fields are null. A real zero stays zero. An empty source array is preserved as an empty array. Missing raw keys remain absent. Do not assume different count fields have identical definitions or that a reported zero proves the underlying metric is available. Instagram may hide or restrict metrics; related visibility flags remain in raw.

Observed source fields include video versions, thumbnails, subtitles, DASH manifest, clips/audio metadata, collaboration and sponsorship metadata, location, preview comments and product tags. Preview comments are not a complete comment export. The local FIELD_INVENTORY.md records all 544 observed nested paths from two public Reels; that sample is not an exhaustive Instagram schema.

Media and profile image URLs can expire. Downloading the media itself, historical snapshots, account discovery, follower exports and private-content access are not included. An author must explicitly be marked public in the response; otherwise no media content is returned.

Failure handling and exports

Successful results go to the default dataset, available as JSON, CSV or Excel. JSON is recommended for nested raw details. Failures go to the key-value store's SUMMARY, leaving successful rows accessible and marking the run failed:

• capture_timeout: no matching fresh full-media response within the capture window.
• privacy_not_verified: public visibility could not be established.
• unsupported_media: the response is not a Reel video.

A timeout does not establish deletion, private visibility or zero plays. Internal errors stop the run without turning them into fake media records. Repeated runs make new observations; resuming the same run avoids re-pushing known results.

Requests from overlapping Actor runs enter a durable FIFO queue on the collector. The collector processes one Reel at a time because it uses one simulator. It accepts up to 100 queued or running requests; when full, the Actor retries the same request ID. The Actor submits each Reel once and polls that request until it finishes. A service restart preserves queued requests. If a restart interrupts an active capture, its outcome is marked unknown and is not collected again, so an ambiguous capture cannot create a duplicate result or charge.

Privacy and pricing

Only the requested media object is captured. Authentication/session fields, tracking tokens and viewer-specific relationship or interaction state are removed recursively. Unknown media fields are retained subject to these rules; the output is sanitized data, not a byte-for-byte raw response.

Pricing is $0.0008 per successfully returned Reel ($0.80 per 1,000 dataset items), through Apify's automatic default-dataset event. There is no per-run start event, and failed observations stay outside the billable dataset. The Actor is configured not to pass Apify platform usage costs through as a separate charge to Actor users.

After publication, use the Actor's issue tab with a run ID for support. Never include credentials or cookies.