Analysis updated 2026-05-18
Translate a foreign-language film's .srt subtitle file into English using a locally running AI model with no data leaving your machine.
Convert subtitle files for Arabic or Persian content where right-to-left text ordering needs to stay correct.
Translate fan-made subtitles for a TV show while supplying context notes so the model uses the right terminology for that show.
| lockhartkz/subllminal | abidoo22/pixelorama-mcp | aditya-pandey/slate | |
|---|---|---|---|
| Stars | 1 | 1 | 1 |
| Language | TypeScript | TypeScript | TypeScript |
| Setup difficulty | moderate | moderate | easy |
| Complexity | 2/5 | 3/5 | 2/5 |
| Audience | general | vibe coder | general |
Figures from each repo's GitHub metadata at analysis time.
Requires a separately running local LLM like Ollama or LM Studio, or a cloud API key, before translating.
SubLLMinal is a desktop application for Windows, macOS, and Linux that translates subtitle files for films and TV shows using an AI language model you run yourself. You open a .srt or .ass subtitle file, pick the source and target languages, connect to a locally running model, and click Translate. The app shows a side-by-side preview of the original and translated text as it works, and saves the finished file when you are done. The local-first design means the AI model runs on your own machine using tools like LM Studio, Ollama, or llama.cpp, so nothing is sent to a third-party server. You can also connect to a cloud API endpoint with your own key if you prefer. Bigger and better-trained models give better results, especially for languages that read right-to-left. The app is careful about what it sends to the model. Timing codes, styling tags, and text formatting are stripped out before any text reaches the model, then put back afterward. This means the model only sees words, not technical subtitle markup, which reduces mistakes and keeps all formatting intact in the output. Lines are sent in small batches with a few already-translated lines included for context, and the app automatically retries or falls back to one-line-at-a-time if a model skips or mangles a line. Right-to-left languages like Arabic and Persian are explicitly supported. A comparison in the README shows that other translation tools can mis-order Latin tokens inside a right-to-left line, and SubLLMinal handles this correctly. For setup, you download the portable executable for your operating system from the releases page and double-click it, no installer is needed. You also need a locally running model or a cloud API endpoint set up separately. macOS users who get a security warning on first launch need to clear it once through System Settings.
A portable desktop app that translates .srt and .ass subtitle files using a locally running AI model, keeping all timing, styling, and right-to-left language support intact.
Mainly TypeScript. The stack also includes TypeScript, Rust, Tauri.
Use freely for any purpose including commercial projects as long as you keep the copyright notice.
Setup difficulty is rated moderate, with roughly 30min to a first successful run.
Mainly general.
This repo across BitVibe Labs
Verify against the repo before relying on details.