# Changelog of Instagram AI Transcript Extractor (`sian.agency/instagram-ai-transcript-extractor`) Actor

- **URL**: https://apify.com/sian.agency/instagram-ai-transcript-extractor/changelog.md
- **Full Actor documentation**: https://apify.com/sian.agency/instagram-ai-transcript-extractor.md

## Changelog — Instagram Transcript Generator Updates

All notable changes to the **Instagram AI Transcript Extractor** (Instagram Transcript Generator) will be documented in this file.

### \[2026-08-28] — v2.2 — Scraping Date & Clickable Dataset Links

#### 🛠️ Improved

- **Scraping Date on Every Row** — every transcript and error row now carries a `scrapingDate` field, so you can filter or group results by the day they were captured without parsing the full processing timestamp.
- **Clickable Dataset Links** — thumbnail, video, and Instagram URL fields now render as clickable links wherever you export or view the dataset, instead of plain text you have to copy by hand.

#### 💎 User Benefits

- Faster filtering: group or sort transcripts by scraping day without touching timestamps
- Cleaner exports: URL fields open directly instead of needing a copy-paste

#### 🎯 Use Cases

- Agencies build a daily digest of new transcripts by grouping rows on the scraping date
- Researchers click straight through to the source reel or thumbnail from the dataset view

### v2.1.35 — Nothing Is Capped Unless You Type A Number (August 2026)

#### 🛑 Changed

- **Channel scrape takes the whole profile by default** — "Amount of Reels to Scrape" defaulted to 10. It is now empty, which means every reel on the profile, inside the date filter if one is set. Type a number to stop there. Free runs are still limited to 5.
- **The comment cap is gone in both directions** — leaving "Max Comments Per Video" empty used to return a single page of comments. It now returns every comment, following the cursor to the end. The hard ceiling of 200 is gone too, so a number you type is the number you get.

### v2.1.33 — Subtitle Files, a Rebuilt Run Report & Honest Errors (August 2026)

#### 🚀 Added

- **Ready-to-use subtitle files** — every result now carries `srtSubtitles` and `vttSubtitles` next to the plain transcript, so an Instagram reel goes straight into a video editor, a YouTube upload or a web player with no conversion step in between.
- **Rebuilt run report** — the Processing Report now opens with the transcripts themselves and a one-click copy button for each, followed by a per-reel table showing creator, detected language, duration and word count.
- **Add-ons you have not tried yet** — the report names the paid extras a run did not use (fast processing, transcript translation, pinned comments), the exact input that switches each one on, and what it currently costs.

#### ⚡ Improved

- **An itemised charges statement** — each run report lists every billable event, how many of each were charged and what they cost, plus what a cheaper Apify plan would have billed for the same run.
- **One-click retry for failed URLs** — anything that did not transcribe is listed with a specific reason and a ready-made retry payload you can paste into a new run. Failed URLs are never charged.
- **Clearer dataset columns** — the Instagram transcript results view now labels every column, so transcripts, engagement metrics and creator fields are easier to scan and export.

#### 🐛 Fixed

- **Honest error messages** — a photo post or image carousel is now named as such instead of being reported as a missing audio track, so an unsupported post type is easy to tell apart from a genuine transcription failure.

#### 💎 User Benefits

- Subtitle-ready Instagram transcripts with no converter in the loop — SRT and WebVTT arrive on the same row as the text
- Every run ends with a plain-language account of what you got and exactly what you paid for it
- Faster recovery from a bad link: the report names the fix and hands back a runnable retry payload

#### 🎯 Use Cases

- Video editors pull SRT subtitles for Instagram reels straight into their timeline without a conversion tool
- Agencies reconcile a bulk Instagram transcription run line by line from the run report itself
- Researchers re-run only the handful of URLs that failed, using the retry payload the report generates

### v2.1.17 — Faster, Leaner Transcription Pipeline (July 2026)

#### ⚡ Improved

- **Faster transcripts** — per-reel processing now typically finishes in 3–10 seconds, in single, bulk, and channel modes alike.
- **Hardened downloads** — a strict time budget on media fetching means a stalled network stream can no longer hang a run.
- **Smarter failure handling** — a reel that is briefly unavailable now returns a clear message and what to do next, instead of stalling the run. Posts that just need another attempt are still recovered for you automatically.
- **Right-sized runs** — default run memory set to 256 MB, with a 4 GB ceiling. Lighter runs start faster.

#### 💎 User Benefits

- Noticeably faster turnaround on every reel, at the same price
- More predictable runs: fewer stalls, cleaner errors on temporarily unavailable posts

### v2.1 — Pinned & Creator Comments (June 2026)

#### 🚀 Added

- **📌 Pinned & Creator Comments** — new "Include Pinned & Creator Comments" option returns the comments a creator pinned to the top of a post (Instagram allows up to 3), plus the creator's own unpinned comments. When the value of a post lives in a pinned comment — the full recipe, ingredient list, or links — you now capture it. Output is a full `comments` array (pinned comments first, then top comments) plus ready-made `pinnedComments` and `creatorComments` shortlists, each with text, author, like count, and pinned/creator flags.

#### 💰 Pricing

- The comments add-on is available on **paid Apify plans**, billed per post **only** when the option is enabled **and** at least one comment is returned. Leave it off for no extra cost.

#### 💎 User Benefits

- Capture recipes, links, and details creators hide in pinned comments — perfect for recipe apps, tutorials, and product research
- Opt-in and pay-as-you-go — transcript-only runs are unaffected

#### 🎯 Use Cases

- Recipe apps extract the full recipe when it's pinned in the comments instead of spoken in the video
- Researchers and marketers capture the source links and details creators add in their own pinned comment

### v2.0 — Transcript Translation into 100+ Languages (June 2026)

#### 🚀 Added

- **🌍 Translate Instagram Transcripts into 100+ Languages** — new "Translate Transcript To…" option delivers a professional-quality `translatedTranscript` in English, German, French, Spanish, Japanese, Chinese, and 100+ more languages (charged per 1,000 transcript characters; a typical reel ≈ 1 unit). The original transcript, segments, and word-level timestamps stay in the source language so timing stays accurate.

#### 💰 Pricing

- **Up to 40% Cheaper Fast Processing** — the fast processing upgrade drops to $0.026 per URL on the free plan and as low as $0.009 per URL on paid plans. Bulk Instagram transcription at speed is now even more affordable.

#### 💎 User Benefits

- Unlock global Instagram content — research international creators and competitors in YOUR language
- One setting, every mode — translation works for single reels, bulk URL lists, and full channel scrapes

#### 🎯 Use Cases

- Marketers monitoring Spanish- and Portuguese-language influencers translate viral reels into English briefs in minutes
- German and French agencies localize English-language viral content for their home markets
- Researchers studying global trends transcribe and translate creator channels across language markets in one run
- Agencies repurposing international viral content get English scripts with word-level timing ready for subtitles

### Speaker Diarization, Language Detection & Enhanced Reel Transcription (April 2026)

#### Added

- **Speaker Diarization** — `speaker` field included in `segments` and `words` arrays (e.g. SPEAKER\_00, SPEAKER\_01). Identify who is speaking in multi-speaker Instagram videos.
- **Language Detection** — `detected_language` field with ISO 639-1 language code. The AI transcription engine now automatically detects the spoken language.
- **Reshare Count** — `reshareCount` engagement metric added to output.
- **Verified Badge** — `isVerified` field indicating whether the creator has a verified account.

#### Improved

- **Enhanced Reel Content Extraction** — Significantly improved Instagram reel transcription, including text extraction from reels that are not directly accessible. Higher success rates and more complete transcript results.

#### Changed

- Updated dataset schema, dataset views, and documentation to include the new output fields.

### Channel Scrape Mode for Bulk Instagram Transcription (January 2026)

#### Added

- **Channel Scrape Mode** — Transcribe entire Instagram channels automatically. Enter a username and get AI transcripts for all reels without collecting URLs manually.
- **Tier-Aware Limits** — FREE tier: 5 reels per run, PAID tier: unlimited reels.
- **Date Filtering** — Use `onlyPostsNewerThan` to only transcribe recent Instagram content.
- **Scheduled Automation** — Set up recurring runs to automatically transcribe latest content from any Instagram creator.

#### Use Cases Enabled

- Competitive intelligence at scale — transcribe competitor Instagram channels in one click
- Content research — analyze trending patterns across entire creator portfolios
- Trend monitoring — schedule weekly/daily Instagram transcription of key accounts

### Word-Level Timestamps & 75% Cheaper Fast Processing (November 2025)

#### Added

- **Word-Level Timestamps** — Get precise timing for every single word in Instagram transcripts. Included for all users at no extra cost.
- **75% Cheaper Fast Processing** — Reduced from $80/1000 to $20/1000 results for bulk Instagram transcription.
- **PAID Tier Speed Boost** — 1-second rate limiting (8x faster than FREE tier).
- **Enhanced Dataset Views** — Optimized field ordering with word-level timing data display.

#### User Benefits

- Process 100+ Instagram reels per hour with word-level timing data
- Professional-grade Instagram video transcripts with precise timestamps for every word
- Faster processing for paid users with optimized rate limiting
- Better value with dramatically cheaper fast processing options