Retro Asset Studio generates images through an AI provider that you connect with your own API key. You configure it per project from Project Settings > API Configuration.
In this alpha, Google Gemini is the recommended and the only enabled provider. Hugging Face and OpenAI are listed but disabled and labelled Coming soon. See AI Providers.
The Provider List¶
Open Project Settings from the sidebar and scroll to the API Configuration section. The card shows three providers:
- Google Gemini (Image Generation) -- Enabled and selected. Calls Google's Gemini API directly with an AI Studio key. Requires a paid Google AI plan for the image generation models.
- Hugging Face (Nano Banana 2) -- Coming soon. Greyed out; cannot be selected in this version.
- OpenAI (gpt-image-1 / DALL-E 3) -- Coming soon. Greyed out; cannot be selected in this version.
Quick Setup Steps (Gemini)¶
- Get a Google AI Studio key at aistudio.google.com/apikey and make sure the linked Google Cloud project has billing enabled.
- In Project Settings > API Configuration, paste the key into the Google AI API Key field.
- Pick an Image Generation Model:
gemini-3-pro-image(recommended default),gemini-2.5-flash-image, or Custom… for another model id. - Click Test Connection to verify the key and model.
- Click Save Settings.
For step-by-step instructions on getting the key and choosing a model, see AI Providers.
Test Connection is billed
Test Connection sends a real generation request to Gemini, which counts against your quota and billing.
Per-Project Configuration¶
Credentials are stored per project. Switching projects automatically switches to that project's API configuration. Deleting a project removes its stored credentials.
Keys are saved locally in the application database, encrypted with Windows Data Protection (DPAPI), so they can only be decrypted by your Windows user account on this PC. They are never included in exported projects. See Privacy & Data.
Troubleshooting¶
Errors appear as notifications in the app's bottom status bar. Each notification has a copy button so you can paste the exact message into a search or a bug report.
| Message | Likely Cause |
|---|---|
| "Please configure your Google AI API key in Project Settings first" | No key has been entered and saved for this project |
| "API key not valid" (400) | The key was pasted incorrectly or deleted in AI Studio |
RESOURCE_EXHAUSTED (429) |
Free-tier key without billing, or a quota hit. Enable billing or wait and retry. |
| "Model not found" (404) | The selected model is not available to your account or region. Pick another model. |
| Generation timeout | High load on the provider. Wait and retry. |
See AI Providers for the full Gemini error table and Troubleshooting for general issues.