Gemini3DS changelog
v2.1 · Aug 10, 2026
📸 Image Prompting
What's new:
- Image prompting: You can now use the 3DS camera to take photos and send them directly to Gemini alongside a custom text prompt.
- Multi-Key Management & Auto-Rotation: Save multiple API keys to your SD card. If a rate limit (HTTP 429) is reached, the app automatically swaps to the next saved key and retries the prompt.
Bug Fixes
- Functional AI Settings: Temperature and Max Tokens settings now correctly map to the Gemini API, granting full control over response creativity and length.
- Mid-Sentence Cutoff Fix: Raised the default Max Token limit from 100 to 1000. Gemini will no longer run out of "thinking tokens" and cut off your answers after a few words.
> 💡 Tip for Camera Users:
> When sending images from the 3DS, stick to the Flash or Flash Lite models. Heavier models take significantly longer to process image data, which can cause the 3DS to hit a network timeout.
v2.0 · Mar 28, 2026
🆕 What's new
- Audio Input Support🎙️: Send voice prompts directly to Gemini using the 3DS internal microphone.
- Request Validation: Added logic to handle audio requests (prevents crashes from clips that are too short or silent).
📥 Installation Guide
3DSX (Homebrew Launcher):
- Download
Gemini3DS.3dsxbelow. - Copy it to the
/3ds/gemini/folder on your SD card. - Launch via the Homebrew Launcher.
v1.0 · Dec 16, 2025
✨ Features
- Text requests
- Settings for: model, temperature and max tokens
- In-app API key management
📥 Installation Guide
3DSX (Homebrew Launcher):
- Download
Gemini3DS.3dsxbelow. - Copy it to the
/3ds/gemini/folder on your SD card. - Launch via the Homebrew Launcher.
