GetTranscribe
AITranscribe social video from Instagram, TikTok, YouTube, Facebook, X, Pinterest, and Google Drive into text, and resolve direct media download links. Agents turn published video into searchable, quotable copy.
What This Integration Enables
Every other transcription connector wants a file. GetTranscribe wants a link to something somebody already published. That difference in intake is the whole product: you point it at an Instagram reel, a TikTok, a YouTube video, a Facebook or X post, a Pinterest video, or a Google Drive link, and it returns the words. No download, no re-upload, no intermediate storage step just to get audio into a transcription service.
Once a job completes, the transcription record carries more than text. It carries word and character counts, duration, the detected platform, author and channel metadata, a generated title, and presigned audio, video, and thumbnail URLs. That turns a scattered set of public posts into something a flow can actually reason over: filter by platform, filter by folder, search the text, compare what an account said in March against what it says now. The connector also separates looking from keeping. Get Video Download URL resolves a direct media link without spending a transcription credit, and Save Video To File persists the media into FlowRunner file storage so it outlives the presigned URL. Those are two different acts, and treating them as one is how teams end up with an archive nobody decided to build.
Without FlowRunner
With FlowRunner
Use Case Scenarios
A weekly sweep of the accounts you watch
A tracked list of creator and competitor accounts gets swept on a schedule. The agent submits each new post URL with Create Transcription Job, polls until completed, and pulls the full record with Get Transcription. It writes the transcript, title, duration, and author metadata into the research base and posts a digest into Slack with what changed in the messaging this week. The team reads positioning shifts as text instead of budgeting two hours to watch reels, and no media is retained unless someone says so.
Your own video turned back into written copy
The posts you published are the easiest content to repurpose and the least likely to get repurposed. A flow takes your own Instagram for Business and YouTube posts, transcribes each one, summarizes the transcript with OpenAI into a blog draft that keeps your own phrasing, and files the draft in WordPress as a draft post for an editor. The video already said the thing. This just stops it from only existing as video.
Prospect research before the call
A named account is added to an outbound sequence. The agent transcribes the prospect's public talks, webinar clips, and product announcement videos, extracts the claims and priorities they stated on the record, and writes a short brief onto the HubSpot record before the first call. The rep walks in knowing what the buyer has already said publicly, in the buyer's own words rather than a summary of a summary.
Human-in-Loop Highlight
The gate here sits on the boundary between reading and keeping. Transcribing a public post is one thing. Save Video To File is another: it resolves the social URL, downloads the media before the presigned link expires, and writes a durable copy of somebody else's video into your storage where it stays. Get Video Download URL does not even spend a transcription credit, which makes hoarding third party media the path of least resistance if nobody puts a stop in the way.
So the agent does not decide retention. It transcribes the sweep, files anything from accounts you own, and stops on the rest, posting to the content lead in Slack: "This week's sweep returned nine third party clips, six from creators and three from competitor accounts. Transcripts are filed. Two are candidates for quoting in the newsletter. Which of these, if any, should I save the media for, and which are transcript only?" The lead answers per clip. That answer is the reason the archive exists, and it is on record when somebody later asks why a competitor's video is sitting in your file storage. Reading in public is free. Keeping a copy is a decision, and this connector makes a person make it.
Agent Capabilities
7 actionsTranscription
5- Create Transcription Job Submit a public video URL and return immediately with a pending job. Accepts Instagram, TikTok, YouTube, Facebook, X/Twitter, Pinterest, and Google Drive links along with direct video files, and takes a model choice, language, prompt, and folder ID. Transcription is asynchronous, so poll until the status is completed.
- Get Transcription Job Retrieve the current state of a job including status, progress messages, the detected platform, and the transcription ID once processing finishes. This is the polling companion to Create Transcription Job.
- List Transcription Jobs Return a paginated list of jobs for the account, newest first, filterable by status. Used to sweep for jobs still processing or to reconcile a batch that partly failed.
- Get Transcription Retrieve a completed transcription in full: transcript text, word and character counts, duration, detected platform, author and channel metadata, a generated title, and presigned audio, video, and thumbnail URLs. The presigned URLs are time limited and expire.
- List Transcriptions Return a paginated list of completed transcriptions, optionally filtered by folder or source platform. The list omits full transcript text and presigned media URLs, so call Get Transcription for the complete record of anything you need in full.
Media
2- Get Video Download URL Resolve a social video URL into a direct, time limited media download URL and thumbnail without spending a transcription credit. The returned URLs expire, so consume them promptly or persist the media instead.
- Save Video To File Resolve a social video URL, download the media before its presigned link expires, and store it in FlowRunner file storage, returning a durable URL later steps can use. Downloaded bytes are validated against known audio and video signatures so an expired or error response is never saved as media. Because this creates a lasting copy of third party content, put it behind an approval rather than running it on every sweep.
Triggers
1 triggersEvent Triggers
1- On Transcription Completed Fires once for every transcription job that reaches the completed state. GetTranscribe exposes no webhook subscription API, so this trigger polls the job list and emits only jobs it has not reported before, at a configurable interval with a thirty second minimum. Use it to start downstream work the moment a transcript exists rather than on a fixed schedule that guesses when jobs will finish.
Frequently Asked Questions
What can FlowRunner do with GetTranscribe?
FlowRunner agents can run Create Transcription Job, Get Transcription Job, and List Transcription Jobs in GetTranscribe, plus 4 more actions.
Does connecting GetTranscribe to FlowRunner require OAuth?
No. GetTranscribe connects to FlowRunner with an API key, no OAuth flow required.
Can GetTranscribe trigger a FlowRunner workflow automatically?
Yes. GetTranscribe supports 1 trigger that can start a FlowRunner workflow automatically.
Start building with GetTranscribe
$100 in credits. No card required. Connect in minutes.