The File Transcribe API Is Live (2026)

The File Transcribe API is for the job the website already does, except a person does not have to click Upload. A Plus account creates a key, your app sends the file, and you poll until the transcript is ready.

It is not a meeting bot, a dialer, or a human transcription desk. If the recording already exists as a file, this is the door into the same speech-to-text, speaker labels, and caption export you get in the editor.

Guide: API documentation. Keys live in dashboard settings. The plan that includes it is Plus.

Who this is for

You want this if some other program already has the recording.

• A course site that stores lesson masters and needs an SRT next to each one.
• A support desk that exports a sampled call and wants the text back in its own tool.
• A script on a server that already downloads the MP4 and should not open a browser.

You do not want this if the work is "open the homepage and fix one interview." That path is still the upload box. The API is the same engine with a key in front of it.

What a call actually does

1. `POST /api/v1/uploads` with the filename and size. You get a signed URL.
2. Put the file on that URL. The bytes do not pass through the API request, so a long lesson can still be a large file.
3. `POST /api/v1/transcriptions` with the upload id, or with a public https link to the file.
4. `GET /api/v1/transcriptions/:id` until `status` is `completed` or `failed`.
5. Download TXT, JSON, SRT, or VTT from the export URL.

The finished item shows up in your library. Open `editor_url` when a name is wrong and you would rather fix it in the browser than in your own code.

Minutes come off the Plus balance. A 40 minute lesson costs 40 minutes, same as an upload from the dashboard. The hourly cap (200 requests per key) is about how often you call the API, not how much audio you can transcribe.

What we left out on purpose

No webhooks yet. Poll. A course batch of 30 lessons is a loop with a sleep, not a callback server.

No Ask AI, no glossary management, no team membership endpoints. Those stay in the product. v1 is: file in, transcript and captions out.

Language codes, errors, and the upload headers are written out at filetranscribe.com/developers. Copy from there. The secret starts with `ft_live_` and is shown once.

Where this replaces a person clicking Upload

A course library is still a sheet of lesson ids, masters, and a spot-check. The API is how an engineer walks that sheet without opening 80 browser tabs. The operating notes are in how to caption an entire course library.

A call center sample is still a sample. Export the WAV, send it, paste the timestamp into the scorecard. The shopping list is transcription software for call centers. The page for the human version of that upload is customer support transcription.

If you are choosing caption software and the output you need is a file, the ranking is best automatic captioning software. The API is that same SRT, requested by code.

Limits worth knowing before you wire it up

Plus is 2 GB and 3 hours per file, about 5,000 minutes a month. A Free or Pro key is rejected. Create the key on the Plus account that should be billed.

Send either an upload id or a public https URL, not both. Links to your own office network are refused. If the file is not on the public internet, use the signed upload.

If you send the same `file_hash` again, you get the existing transcript back unless you pass `force: true`. That is there so a retry does not burn the balance twice.

SRT and VTT follow the paid export you already have in the editor. JSON and TXT are there for apps that want the words, not a caption file.

FAQ

Do I need a separate API plan?

No. Plus includes it. Pro stays the website: editor, captions, summary, translation. If an engineer is embedding this in another product, that account should be Plus.

Does the file stay in File Transcribe?

Yes. Retention matches the Plus plan (90 days). Delete through the API or from the library when you are done with it.

Can I get a webhook when it finishes?

Not in this version. Poll the transcription id. Webhooks are the next thing people will ask for, and they are not shipping in v1.

Where do I get a key?

Sign in on a Plus account and open API keys. Read the introduction before you send the first file.

Create the key, send one short recording, and open the result in the editor before you point a whole library at it. Pricing is where Plus starts.

More guides

• Detect topics and keywords with AI
• AI sentiment and intent in transcriptions
• How AI transcriptions save time
• Test transcription accuracy
• Transcription guides

Further reading

• Google Speech-to-Text docs
• Zoom cloud recording help