- What is SubRay?
- Key Features
- Comparison with Subtitle Edit
- Download and Installation
- Detailed Tool Guide
- How to Use
- About
SubRay is not just an ordinary translation tool. It's a comprehensive and advanced tool for translating subtitle files (SRT, ASS, VTT) using the latest generative AI technologies, delivering accurate translations with smart token optimization and professional results.
- 4 Built-in Tools
- Lowest Token Consumption
- Accurate Cost Estimation
- Direct Processing
- Simplified Interface
No need to write any translation rules! The tool comes with a complete translation system that understands:
- Preserving line count (each line translates to one line)
- Keeping formatting codes (ASS/SSA codes) in their exact positions
- Accurate translation without censorship or meaning changes
- Using formal/standard language
Custom Rules: Are only for your personal additions if you want, such as:
- "Use informal/casual tone instead of formal"
- "Keep character names untranslated"
- "Add honorifics like Mr./Mrs. before names"
| Tool | Description | Special Features |
|---|---|---|
| Manual Batching | Extract texts and re-merge them | • Smart line splitting • All formats supported • Safe file saving |
| Smart Rename | Smart renaming and sorting | • Video sync • Custom naming • Archive support |
| LM Studio Client | Local translation using GPU | • 100% Free • Up to 4 machines • Smart load balancing |
| External API | Cloud-based translation | • 6 different providers • 50+ AI models • Instant cost estimation |
| Provider | Supported Models | Special Features |
|---|---|---|
| OpenAI | GPT-5.2, GPT-5.1, GPT-4.1, GPT-5-mini | High linguistic accuracy |
| Anthropic | Claude 4.5 Sonnet, Claude 4 Sonnet | Deep contextual understanding |
| Gemini 3 Pro, Gemini 3 flash | High speed and efficiency | |
| DeepL | Standard (Free/Paid) | XML Mode support, timing code handling |
| OpenRouter | 1000+ models | Open model marketplace |
| xAI (Grok) | Grok 4.1, Grok 4 fast | Economic alternative |
- Full LM Studio support
- Processing with up to 4 simultaneous machines
- Smart load balancing
- 100% Free translation
- Optimized processing speed
- Comprehensive format support: SRT, ASS, VTT, SSA
- Direct archive processing: ZIP, RAR, 7z (no manual extraction needed)
- Smart code and format extraction
- Precise re-merging while preserving timing
- Smart queue system
- Multi-threading processing
- Automatic retry (Exponential Backoff)
- Optimized memory management
- Accurate cost calculator before starting
- Detailed statistics: tokens, characters, requests
- Usage tracking per provider
| Feature | Subtitle Edit | SubRay |
|---|---|---|
| Multiple Files | Single file only | Unlimited queue |
| Custom Rules | Not available | Full control |
| Archive Support | Requires manual extraction | Direct processing |
| Cost Estimation | Billing surprises | Accurate pre-calculation |
| Retry System | Final failure | Advanced smart system |
| Local Translation | Single machine only | 4 machines |
| User Interface | Complex | Easy and simple |
-
Download the Tool:
- Go to Latest Release
- Download SubRay_vx.x.x_Setup.exe
-
Run the Installer:
- Download the compressed file
- Extract and run
SubRay_Setup.exe - Follow the installation wizard
- Operating System: Windows 10/11 (64-bit)
- Memory: 4GB RAM minimum (8GB recommended)
- Storage: 500MB free space
- Graphics Card: 8GB VRAM or higher (for LM Studio usage)
- Internet: Required for cloud translation (optional)
⚠️ Windows SmartScreen Warning: When running the installer for the first time, Windows SmartScreen may display a warning because the application is new and not yet digitally signed. This is normal for new open-source software. Click "More info" then "Run anyway" to proceed with installation.
What is it? A dedicated window for extracting texts from subtitle files and re-merging them after external translation. Very useful when you want full control over the translation process.
When to use it?
- When you want to translate files in batches for external models like ChatGPT Web
- You want to split files for easier handling and lower context consumption
- If you want to edit or translate specific lines without dealing with timing and formatting
- When working with models that require short texts
- For review and proofreading before final merging
How it works in detail:
- Smart Extraction: Converts subtitle file into text files split by the specified number of lines
- Timing Preservation: Keeps all timing codes and original formatting
- External Translation: Transfer extracted texts to an external service (ChatGPT, Claude, etc.)
- Precise Re-merging: Merges translated texts with the original file while preserving timing
Special Features:
- Smart line splitting (Normal or Smart Chunks)
- 100% accurate timing preservation
- All format processing (SRT, ASS, VTT)
- Safe file saving without data loss
- Manual text editing before merging
- Direct compressed file support
What is it? A window for local translation using AI models installed on your machine. Provides complete privacy and free translation.
When to use it?
- You want 100% free translation without limits
- You have a powerful machine with a good GPU
- You want complete privacy (no files sent to external servers)
- You have multiple machines to distribute the load
- You work in an offline environment
How it works in detail:
- LM Studio Setup: Download and run a local model on your machine
- Connection: Link SubRay to LM Studio via local API
- Translation: Use the local model for translation without internet
- Smart Balancing: Distribute lines across 4 machines if available
Special Features:
- Load balancing across 4 machines on local network
- Various local models (Llama, Mistral, Aya, and more)
- Local API key storage without risks
- Custom translation rules support
- Instant translation without waiting
- Complete privacy protection
What is it? A window for cloud translation using paid AI services. Offers the highest translation quality with a wide variety of models.
When to use it?
- You want the highest possible translation quality
- You have an API key from a trusted service provider
- You want a wide variety of models (50+ models)
- You want accurate cost estimation before starting
- Working on professional projects requiring high accuracy
- You need support for rare languages or special dialects
How it works in detail:
- Provider Selection: Choose from 6 different providers (OpenAI, Google, Anthropic, DeepL, OpenRouter, xAI)
- Model Setup: Select from 50+ available models with price preview
- Cost Estimation: Accurate cost calculation before starting with token and character details
- Translation: Use API for fast and accurate translation with retry system
Special Features:
- 6 different providers to choose from
- 50+ advanced AI models (GPT-5.2, Claude 4, Gemini 3 Pro, and more)
- Instant cost estimation with full details
- Complex and custom translation rules support
- Automatic retry system
- Detailed usage and cost statistics
What is it? A window for automatically reordering and renaming subtitle files to match video names. Useful for organizing large libraries of series and movies.
When to use it?
- You have a folder of unorganized videos and subtitles
- You want to sync subtitle names with episode names
- You want to apply a unified naming pattern to a large collection
- You want to reorder many files at once
- Working on a large archive of series
- You want to export organized files in a specific format
How it works in detail:
- File Scanning: Searches for video and subtitle files in the specified folder
- Smart Matching: Matches episode numbers between video and subtitle using smart algorithms
- Pattern Application: Applies the specified naming pattern to each file
- Export: Save as ZIP or directly in folder while preserving structure
Special Features:
- Smart automatic sync with videos
- Complex and customizable naming patterns support
- Direct archive processing (ZIP, RAR, 7z)
- Comprehensive preview before application to avoid errors
- Different patterns for series and movies
- Export in various formats while maintaining quality
The tool has 4 main tabs, each with a specific use:
| Tab | Purpose | Usage Steps |
|---|---|---|
| Manual Batching | Extract texts and re-merge after external translation | 1. Drag subtitle files or click Browse 2. Set number of lines per text file 3. Click Extract to extract texts 4. Translate text files externally 5. Click Merge to re-merge |
| LM Studio Client | Free local translation using GPU | 1. Install LM Studio from lmstudio.ai 2. Load AI model and start local server 3. Enter server address and click Connect 4. Drag subtitle files 5. Select target language and click Start |
| External API | High-quality cloud translation | 1. Enter API Key for chosen provider 2. Select model and target language 3. Drag subtitle files 4. Click Check Cost to see the cost 5. Click Start Translation |
| Smart Rename | Rename subtitles to match videos | 1. Select folder containing videos and subtitles 2. Click Scan to scan files 3. Review preview and suggested names 4. Modify naming pattern if desired 5. Click Rename or Export |
| Step | Action |
|---|---|
| 1. Installation | Download LM Studio from lmstudio.ai and install it on your machine |
| 2. Load Model | Open LM Studio and choose a suitable model like aya-expanse-8b-GGUF or mistral-8b-instruct |
| 3. Start Server | Go to Local Server and click Start Server |
| 4. Connect | In SubRay, go to LM Studio Client tab and enter server address |
| 5. Connection | Click Connect and verify connected status appears |
SubRay is a free project designed to help translators translate subtitle files using AI
You can support the developer by:
- Giving a positive rating ⭐
- Suggesting new features
- Reporting bugs for improvement
Report a Bug • Suggest an Idea
Free License for Use - See LICENSE for full terms and conditions
License Covers:
- Free personal, educational, and commercial use
- Installation on multiple devices for same user
- Creating backup copies
Main Restrictions:
- Distribution or sale of executable not allowed
- Reverse engineering or modification not allowed
- All copyrights reserved to the developer


