Whether you need show notes, a searchable archive, or subtitles for a YouTube upload, transcribing your podcast is one of the highest-ROI things you can do for SEO. And in 2026, you don't need to pay $1.50/minute to Rev or wait two hours. Here's the fastest free way.
ffmpeg -i episode.wav -b:a 96k episode.mp3 to compress a WAV to MP3 in seconds. A 30-minute episode becomes ~21MB.From your recording software (Descript, GarageBand, Audacity, Adobe Audition), export the episode as MP3 at 128kbps or lower. Mono is fine — you don't need stereo for transcription.
Go to ScribeForge.io. No login, no email address, no credit card. The tool is in your browser immediately.
The grey dashed area at the top accepts drag-and-drop. Alternatively, click it to open a file picker. You'll see the filename and size confirm once it's loaded.
ScribeForge sends the file to xAI's Grok STT API and returns the transcript. For a 30-minute episode, expect 8–15 seconds of processing time.
Click Copy to put the full transcript on your clipboard, or Download .txt to save it. The timestamps panel shows you when each segment was spoken — useful for chapter markers.
Paste the transcript into a new page or post on your website. Add the episode title, guest name, and publication date as structured data. Google will start indexing the text within days, and you'll begin ranking for questions your guest answered in the episode.
Feed the transcript into any LLM (ChatGPT, Claude, Grok) with a prompt like: "Summarize this podcast transcript into 5 bullet-point show notes and extract the 3 most quotable moments."
The downloaded .txt file can be converted to .srt format using free tools like subtitletools.com. The segment timestamps from ScribeForge give you the timing you need.
Pick the most interesting 2–3 minutes of the episode, copy that segment's text from the transcript, lightly edit for readability, and paste into your newsletter as a "quote of the week."
| Tool | Price | Account required | Quality |
|---|---|---|---|
| ScribeForge | Free (2/day) or $9/50 | No | Grok STT — very high |
| Otter.ai | $16.99/mo | Yes | Good |
| Descript | $24/mo | Yes | Good + editor |
| Rev (AI) | $0.25/min | Yes | Good |
| Rev (Human) | $1.50/min | Yes | Excellent |
| Whisper (self-hosted) | Free | No | High (setup required) |
The file size limit is 25MB. A 2-hour MP3 at 128kbps is about 115MB — too large. Split the episode into 30-minute segments with Audacity (File → Export → Export Multiple), transcribe each segment, then combine the texts.
Yes — Grok STT provides segment-level timestamps and speaker labels where it can identify distinct voices. For a two-host format, the transcript will be segmented by speaker turns.
Grok STT supports 50+ languages and detects the language automatically. If your podcast is in Spanish, French, German, Italian, Portuguese, or most major world languages, it will transcribe correctly without any setting change.
Transcribe your next podcast episode for free — no account, no email, no credit card.
Try it now →