good-ttvideo2text_skill

Overview

This skill extracts audio from short Douyin/TikTok videos and transcribes speech to text with per-sentence timestamps. It supports both a web UI for interactive use and a command-line script for automation, returning results as readable text or structured JSON. Use it when you need fast, timestamped transcriptions from short social videos.

How this skill works

Provide a Douyin/TikTok URL or paste share text and the skill extracts the real video link, downloads or references the audio, and submits it to an ASR service. The ASR returns text and timing, and the tool formats results into full transcription, sentence-level timestamps, and optional JSON metadata. Web UI shows progress and cookie management; CLI supports batch and scripted workflows.

When to use it

• You have a Douyin/TikTok short-video URL and need a readable transcript with timestamps.
• You want to automate transcription within an AI workflow or batch process via command line.
• You need JSON output for downstream parsing, search, or indexing.
• You must transcribe videos that require login cookies (with configured cookies).
• You want quick human-readable results via a local web UI with live progress.

Best practices

• Provide the exact share text or direct video URL to ensure correct extraction.
• Configure platform ASR environment variables before running automated jobs.
• Add login cookies in settings for restricted videos to avoid 'video not found' errors.
• Use JSON output for programmatic consumption and text output for human review.
• Retry with the web UI when ASR tasks time out to monitor progress interactively.

Example use cases

• Journalists extracting quotes and timestamps from a viral TikTok clip for an article.
• Researchers building a searchable corpus of short-video speech using JSON output.
• Content creators repurposing spoken content into captions, blog posts, or summaries.
• Customer support or moderation teams auditing short-video audio for policy review.
• Automation pipelines that transcribe batches of share-text links for indexing.