Translate SRT subtitle files using the power of Google Gemini AI! π
Gemini SRT Translator is a powerful python tool to translate subtitle files using the power of Google Gemini AI. Perfect for anyone needing fast, accurate, and customizable translations for videos, movies, and series.
- π€ SRT Translation: Translate
.srt
subtitle files to a wide range of languages supported by Google Gemini AI. - β±οΈ Timing & Format: Ensures that the translated subtitles maintain the exact timestamps and basic SRT formatting of the original file.
- πΎ Quick Resume: Easily resume interrupted translations from where you left off.
- π§ Advanced AI: Leverages thinking and reasoning capabilities for more contextually accurate translations (available on Gemini 2.5 models).
- π₯οΈ CLI Support: Full command-line interface for easy automation and scripting.
- βοΈ Customizable: Tune model parameters, adjust batch size, and access other advanced settings.
- ποΈ SRT Extraction: Extract and translate SRT subtitles from video files automatically (requires FFmpeg).
- π΅ Audio Context: Extract audio from a video file or provide your own to improve translation accuracy (requires FFmpeg).
- π Description Support: Add a description to your translation job to guide the AI in using specific terminology or context.
- π List Models: Easily list all currently available Gemini models to choose the best fit for your needs.
- π Auto-Update: Keep the tool updated with automatic version checking and update prompts.
- π Logging: Optional saving of progress and 'thinking' process logs for review.
pip install --upgrade gemini-srt-translator
It's best practice to use a virtual environment. This is especially recommended as gemini-srt-translator installs several dependencies that could potentially conflict with your existing packages:
# Create a virtual environment
python -m venv venv
# Activate the virtual environment
# On Windows:
venv\Scripts\activate
# On macOS/Linux:
source venv/bin/activate
# Install inside the virtual environment
pip install --upgrade gemini-srt-translator
- Go to Google AI Studio.
- Sign in with your Google account.
- Click on Generate API Key.
- Copy and keep your key safe.
You can provide your API key in several ways:
- Environment Variable (Recommended): Set the
GEMINI_API_KEY
environment variable. This is the most secure and recommended method.
-
macOS/Linux:
export GEMINI_API_KEY="your_api_key_here" export GEMINI_API_KEY2="your_second_api_key_here"
-
Windows (Command Prompt):
set GEMINI_API_KEY=your_api_key_here set GEMINI_API_KEY2=your_second_api_key_here
-
Windows (PowerShell):
$env:GEMINI_API_KEY="your_api_key_here" $env:GEMINI_API_KEY2="your_second_api_key_here" # Optional for additional quota
-
Command Line Argument: Use the
-k
or--api-key
flaggst translate -i subtitle.srt -l French -k YOUR_API_KEY
-
Interactive Prompt: The tool will prompt you if no key is found
gst translate -i subtitle.srt -l French
-
Python API: Set the
gemini_api_key
variable in your scriptimport gemini_srt_translator as gst gst.gemini_api_key = "your_api_key_here"
# Using environment variable (recommended)
export GEMINI_API_KEY="your_api_key_here"
gst translate -i subtitle.srt -l French
# Using command line argument
gst translate -i subtitle.srt -l French -k YOUR_API_KEY
# Set output file name
gst translate -i subtitle.srt -l French -o translated_subtitle.srt
# Extract subtitles from video and translate (requires FFmpeg)
gst translate -v movie.mp4 -l Spanish
# Extract and use audio from video for context (requires FFmpeg)
gst translate -v movie.mp4 -l Spanish --extract-audio
# Interactive model selection
gst translate -i subtitle.srt -l "Brazilian Portuguese" --interactive
# Resume translation from a specific line
gst translate -i subtitle.srt -l French --start-line 20
# Suppress output
gst translate -i subtitle.srt -l French --quiet
# Full-featured translation with custom settings
gst translate \
-i input.srt \
-v video.mp4 \
-l French \
-k YOUR_API_KEY \
-k2 YOUR_SECOND_API_KEY \
-o output_french.srt \
--model gemini-2.5-pro \
--batch-size 150 \
--temperature 0.7 \
--description "Medical TV series, use medical terminology" \
--progress-log \
--thoughts-log \
--extract-audio \
# Show all available commands and options
gst --help
# Show specific command help
gst translate --help
import gemini_srt_translator as gst
gst.gemini_api_key = "your_api_key_here"
gst.target_language = "French"
gst.input_file = "subtitle.srt"
gst.translate()
Just run again with the same parameters, or specify the start line:
import gemini_srt_translator as gst
gst.gemini_api_key = "your_api_key_here"
gst.target_language = "French"
gst.input_file = "subtitle.srt"
gst.start_line = 20
gst.translate()
gemini_api_key2
: Second key for more quota (useful for free Pro models).video_file
: Path to a video file to extract subtitles and/or audio for context (requires FFmpeg).audio_file
: Path to an audio file to use as context for translation (requires FFmpeg).extract_audio
: Whether to extract and use audio context from the video file (default: False).output_file
: Name of the translated file.start_line
: Starting line for translation.description
: Description of the translation job.batch_size
: Batch size (default: 300).free_quota
: Signal GST that you are using a free quota (default: True).skip_upgrade
: Skip version upgrade check (default: False).use_colors
: Activate colors in terminal (default: True).progress_log
: Enable progress logging to a file (default: False).thoughts_log
: Enable logging of the 'thinking' process to a file (default: False).quiet_mode
: Suppress all output (default: False).resume
: Skip prompt and set automatic resume mode.
model_name
: Gemini model (default: "gemini-2.5-flash-preview-05-20").temperature
: Controls randomness in output. Lower for more deterministic, higher for more creative (range: 0.0-2.0).top_p
: Nucleus sampling parameter (range: 0.0-1.0).top_k
: Top-k sampling parameter (range: >=0).streaming
: Enable streamed responses (default: True).- Set to
False
for bad internet connections or when using slower models.
- Set to
thinking
: Enable thinking capability for potentially more accurate translations (default: True).- Only available for Gemini 2.5 models.
thinking_budget
: Token budget for the thinking process (default: 2048, range: 0-24576, 0 also disables thinking).- Only available for Gemini 2.5 Flash (for now).
import gemini_srt_translator as gst
gst.gemini_api_key = "your_api_key_here"
gst.gemini_api_key2 = "your_api_key2_here"
gst.target_language = "French"
gst.input_file = "subtitle.srt"
gst.output_file = "subtitle_translated.srt"
gst.video_file = "video.mp4"
gst.audio_file = "audio.mp3"
gst.extract_audio = False
gst.start_line = 20
gst.description = "Medical TV series, use medical terms"
gst.model_name = "gemini-2.5-pro-preview-03-25"
gst.batch_size = 150
gst.streaming = True
gst.thinking = True
gst.thinking_budget = 4096
gst.temperature = 0.7
gst.top_p = 0.95
gst.top_k = 20
gst.free_quota = False
gst.skip_upgrade = True
gst.use_colors = False
gst.progress_log = True
gst.thoughts_log = True
gst.quiet_mode = False
gst.resume = True
gst.translate()
gst list-models -k YOUR_API_KEY
import gemini_srt_translator as gst
gst.gemini_api_key = "your_api_key_here"
gst.listmodels()
If you prefer a user-friendly graphical interface over command-line usage, be sure to check out:
- π Gemini SRT Translator GUI (by @mkaflowski)
- π Gemini SRT Translator GUI (by @dane-9)
Perfect for users who want the same powerful translation capabilities with an intuitive visual interface!
Distributed under the MIT License. See the LICENSE file for details.
Thank you to all who have contributed to this project:
- MaKTaiL - Creator and maintainer
- CevreMuhendisi
- angelitto2005
- sjiampojamarn
- mkaflowski
Special thanks to all users who have reported issues, suggested features, and helped improve the project.