Note: Remember that you have to download each update manually. GameTranslate does not currently have any auto-update support!

Changelog

• Added main app localisation for every language that the internal translator supports (you can change this in the General tab)
• Added 'User Config Presets' so that users now can save many different configurations and switch between them with a click
• Improved the translation tutorial (unfortunately... it still sucks.. but I plan to expand on it later)
• Added 'Default capture mode' user config to General settings
• Moved 'MangaOCR' straight into 'RapidOCR' (it was just confusing for users)  If you want to read Japanese Manga, set Reading direction to Right to left and Writing Orientation to Vertical in the OCR tab. If you're reading non-japanese manga/comics, just check the 'Comic mode' checkbox as before in the OCR tab.
• Romanization (Romaji) can now be used no matter if the source or target is Japanese
• Romanization (Romaji) can now be used solely without doing any translation. Just pick Japanese as source, set your romanization hotkey and enjoy practising!
• Added traditional Chinese model
• Added English to Icelandic model
• Updated BG-EN, CA-EN, CS-EN, DE-EN, EN-JA, EN-UK, EN-ZH, ES-EN, ET-EN, FI-EN, FR-EN, IS-EN, IT-EN, NL-EN, PT-EN offline models (Language codes) (You will have to delete and redownload to update these)
• Fixed rare exit crash in Desktop/Attached mode
• Fixed Desktop/Tutorial not shutting down properly on some PCs
• Launching Desktop/Attached when an instance is already running will no longer block the new launch, but rather force exit the first instance and start a new one
• Capture mode hotkey default is now F8
• Capture mode hotkey no longer hinders input to fall through to other applications in Desktop/Attached mode
• Added confirmation dialog to 'Reset to default' button
• Y threshold default in OCR tab now defaults to 0.8
• As a first move to completely get rid of Tesseract -- All languages in Internal/DeepL that support RapidOCR will now be unable to select Tesseract. This is also an attempt to remove 'user errors' related to selecting language/OCR models.
• + a ton of other minor improvements