julien-media-subtitle-translation

SRT subtitle translation converts subtitle files using LLM APIs while preserving formatting and HTML tags. Prepare your SRT files, specify source and target languages, configure your LLM provider (OpenRouter, OpenAI, Ollama, or Hostinger), and run the translator script to output translated files with restored formatting.

Batch translation processes multiple SRT files efficiently in a single workflow. The Skill handles sequential or parallel processing of subtitle files, applies consistent language pair rules across episodes, manages retries for failed translations, and outputs all translated files with cost estimation for your entire batch.

HTML tag handling extracts markup before translation and restores it in the output, ensuring tags like `<i>`, `<b>`, and timing codes remain valid and correctly positioned in translated subtitles without breaking SRT structure or player compatibility.

Supported providers include OpenRouter, OpenAI, Ollama (local or Hostinger-hosted). The Skill calculates per-file and batch-level costs based on your chosen model's pricing, token counts, and language pair complexity, helping you plan budgets before running large translation jobs.

Retry logic automatically re-attempts failed translations with exponential backoff, logs errors for debugging, and ensures partial batches can resume without re-translating completed files, improving reliability for large-scale or unstable provider connections.

The Skill automatically parses SRT formatting—extracting timecodes, sequence numbers, and subtitle text—then restores the exact SRT structure after translation, so standard, well-formed SRT files work without preprocessing.