- Home
- Skills
- Translation & Localization
- txtify transcribe
Works with the AI Agents you already use
txtify transcribe
Stop sending private media to cloud transcription services.
Free
txtify transcribe
Example session with this skill installed
Make English subtitles for https://www.youtube.com/watch?v=jNQXAC9IVRw with my local Txtify and save them in ~/Videos/subs.
- Read your context and instructions
- Compiled the txtify transcribe
Txtify is running on http://127.0.0.1:8011 (health check ok). Before starting I confirmed with you:
- Source: YouTube video
jNQXAC9IVRw("Me at the zoo", 19 s), downloaded by your Txtify container - Model:
whisper_tiny, languageen, no translation - Save to:
~/Videos/subs(exists and is writable)
| Step | Result |
|---|---|
| Submit | {"message":"Transcription started successfully.","pid":4} |
| Final status | Completed successfully! |
| Time | 7.48 s |
| Saved | ~/Videos/subs/me-at-the-zoo.srt (new file, nothing overwritten) |
me-at-the-zoo.srt
1
00:00:00,920 --> 00:00:05,700
Alright so here we are one of the elephant's cool thing
2
00:00:05,700 --> 00:00:10,290
for these guys it's that they have really really long
3
00:00:12,200 --> 00:00:18,380
and that's cool and that's pretty much all it is to say
Next steps
whisper_tinyis fast but drops words. For anything you'll publish, run it again withwhisper_baseor bigger.- Want the
.vttor.txttoo, or the whole job as a zip? Same job, no re-transcription needed.
Connects securely to your tools. The creator never sees your data.
About this skill
Stop sending private media to cloud transcription services. This skill allows you to transcribe audio, video, and YouTube links locally using Txtify, an open-source Docker container powered by OpenAI's Whisper models. By running transcription on your own machine, you maintain full control over your data while generating high-quality text or subtitle files.
What it does
- Local transcription converts audio and video files into text without cloud uploads.
- YouTube processing downloads and transcribes videos directly from a URL.
- Subtitle generation exports files in
srt,vtt,sbv, andpdfformats. - Model selection allows choosing between Whisper
tiny,base,small,medium, orlargefor speed or accuracy. - Optional translation integrates with DeepL to translate transcripts into multiple languages.
How it works
- Environment check probes for a running Txtify instance on
localhost:8011or offers to start the Docker container. - Job configuration collects the media source, preferred Whisper model, and target output formats.
- Submission uploads the local file or provides the YouTube URL to the Txtify instance via curl.
- Monitoring tracks the transcription progress (0-100%) and alerts you when the job reaches the export phase.
- Safe export saves the requested formats to an approved local folder using atomic write operations to prevent overwrites.
Frameworks & tools
This skill utilizes docker for container management, curl for API communication, and Whisper-based stable-ts for speech-to-text. It supports mp3, mp4, m4a, wav, webm, ogg, flac, aac, and mov file types.
Why this beats prompting it yourself
Writing manual curl commands for multi-step multipart uploads and polling loops is error-prone. This skill automates the entire lifecycle, from container health checks to safe file handling, ensuring you don't accidentally overwrite data or lose track of long-running jobs.
Use cases
- Generate
.srtsubtitles for a long-form video interview locally. - Create a text transcript of a podcast recording for blog content.
- Transcribe a YouTube lecture without using a paid cloud service.
- Translate a foreign language recording into English via DeepL integration.
Known limitations
Does not support speaker diarization, microphone input, or batch processing of folders. Large Whisper models require up to 10 GB of RAM allocated to Docker.
No speaker diarization or live mic support. YouTube downloads are subject to rate limits. v1.1 image has a 15-minute video limit. Large models require significant RAM.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before any payment.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Free forever
No account required to browse
Trust & safety
Security scan
Security scan passed 20 days ago
- Free with an Agensi account
Requires
Permissions
Allowed Hosts
Terminal: runs curl against your local Txtify (127.0.0.1:8011) and, only after you agree, docker pull/run. Read files: the media file you choose, uploaded only to your own container. Write files: only into the folder you approve; it never overwrites and never follows symlinks. Network: Docker Hub for the one-time image pull; from inside the container, the Whisper model download, YouTube (URL jobs only) and DeepL (only if you turn translation on). No environment variables are read, and nothing is sent to Agensi or the author.