MaKTaiL/gemini-srt-translator

What it solves

Gemini SRT Translator is a tool designed to translate subtitle files (.srt and .ass) using Google Gemini AI. It addresses the challenge of maintaining precise timestamps and formatting while ensuring translations are contextually accurate, especially for videos, movies, and series.

How it works

The tool leverages the Google Gemini API to process subtitles in batches. It can use text-based context (previous lines) and multimodal audio context (extracting audio from video files via FFmpeg) to improve translation quality. It supports both a command-line interface (CLI) and a Python API. For advanced users, it provides a SubtitleSession engine that handles parsing, batching, and progress tracking, allowing the tool to be integrated into other AI workflows or used with different LLMs.

Who it’s for

Content creators, translators, and movie/series enthusiasts who need fast, accurate, and customizable subtitle translations or transcriptions of audio/video files.

Highlights

• Multimodal Context: Extracts audio from video files to provide the AI with audio context for better accuracy.
• Transcription: Directly transcribes audio or video files into subtitle files.
• Format Preservation: Maintains exact timestamps and native SRT/ASS formatting.
• Agent Integration: Can be installed as a skill for AI coding agents like Claude Code, Cursor, and Google Antigravity.
• Resume Capability: Allows users to resume interrupted translation jobs from a specific line.
• Customizable AI Settings: Supports tuning model parameters (temperature, top_p, top_k) and reasoning capabilities (thinking levels) for Gemini 2.5 and 3 models.