Txtify skill
Skill turning audio, video and YouTube links into transcripts and subtitles via your own local Whisper Txtify in Docker.
- 136 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- The skill runs `curl` and `docker` commands and reads the files you pick.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/txtifyDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill runs `curl` and `docker` commands and reads the files you pick.
This is third-party code. Review the repository files before installing.
What it does
The `txtify-transcribe` skill drives your own Txtify instance, a Whisper-based web app, and returns transcripts as `txt`, `srt`, `vtt`, `sbv` and `pdf`. The source is a local file or a single YouTube link, speech language and model size are selectable, and DeepL translation is optional and needs your own key. The agent first looks for a running Txtify on `127.0.0.1:8011` and checks `/health`. If it is absent, it offers to pull the `lkmeta/txtify` image and start a container, but only after you agree. Transcription runs locally and files do not go to a cloud service. Results are saved to a folder you choose.
Who it is for. Editors, journalists and developers who need text or subtitles from recordings without uploading files to cloud services.
Good fit when
- An interview or lecture recording must become a text file.
- A video needs `srt` or `vtt` subtitles.
- A YouTube video must be transcribed without sending it to a third-party service.
- A transcript must be translated with your own DeepL key.
Not a fit when
- You need to tell who is speaking; speaker labels are not supported.
- You need live microphone transcription.
- You want a whole folder or playlist processed in one run.
Example request
Transcribe interview.mp3 in English with the small model and save srt subtitles to the transcripts folder.Limitations
Txtify has no documented public API: the skill calls the same endpoints as the web page and falls back to the browser if a response is unexpected. The service has no authentication, so it must be bound to `localhost` only. The image weighs 2–3 GB and Whisper weights download on first use of each model, from 75 MB to 3 GB. The skill does not do speaker diarization, playlists or live transcription.
How to disable. Stop and remove the container with `docker stop txtify` and `docker rm txtify`, then delete the `txtify-transcribe` skill folder. The `txtify-output` and `txtify-cache` volumes are removed separately.
FAQ
Where do my files go?
The local container does the transcription. Only the YouTube download for link jobs and the transcript text sent to DeepL, if translation is on, leave the machine.
Which formats are available?
`txt`, `srt`, `vtt`, `sbv` and `pdf`.
Why bind the service to localhost only?
Txtify has no authentication and every job is readable by its number. So the container runs with `-p 127.0.0.1:8011:8011`.
What downloads on first run?
A 2–3 GB Docker image and the weights of the chosen Whisper model, from 75 MB for tiny to about 3 GB for large.
Is translation available?
Yes, via DeepL, but only with your key in the `.env` file. Without a key translation stays off.
Discussion
No questions or reviews yet.
Related
SkillsA Codex and Claude Code skill: subtitles and a summary from a Douyin, Xiaohongshu, Bilibili or YouTube link or a local file
The skill turns a video or podcast link or file into a transcript and summary using platform subtitles and Whisper.
A skill that turns a Bilibili, Douyin or YouTube video link into timestamped notes through local transcription
Whisper is a CLI for audio transcription, with model sizes for different hardware and output in the original spoken language.
A skill and scripts for transcribing audio through Yandex SpeechKit: audio preparation, Object Storage upload and asynchronous v3 recognition
watch-cli is a skill and CLI that hands an agent the frames and transcript of any social video so it can read the clip itself.
README in short
The repository holds the Txtify web app itself and the `txtify-transcribe` agent skill under `skills/`, with end-to-end test scripts for it.