Choose and configure an AI Provider
Choose where AI processing runs, connect a supported Provider, test the connection, and activate the configuration you want RenamerX to use.
RenamerX supports six AI Provider types. Custom Provider also lets you connect to other OpenAI-compatible AI Providers.
RenamerX Built-in Local AI is the recommended starting point. RenamerX manages the local AI runtime for you, processes files on your device, and does not require an API key or a separate AI service. Choose another AI Provider when you already use a compatible local service such as Ollama or LM Studio, or when you want to use a cloud model. Before connecting one, check where it runs, whether the model can understand images, and which service terms or usage charges apply.
Choose a Provider
For Ollama, LM Studio, OpenAI, Google Gemini, and Custom Provider, choose a model that supports image input. RenamerX needs a vision-capable model to understand supported images, scanned PDFs, and videos.
| Provider | Where AI processing happens | API key | What you configure |
|---|---|---|---|
| RenamerX Built-in Local AI | On this device | No | RenamerX manages the model, engine, and endpoint. |
| Ollama | At the server URL you enter | No | Server URL and a compatible vision model. The default URL is http://127.0.0.1:11434/v1/. |
| LM Studio | At the server URL you enter | Optional | Server URL and a compatible vision model. The default URL is http://127.0.0.1:1234/v1/. |
| OpenAI | OpenAI cloud | Required | API key and model. |
| Google Gemini | Google cloud | Required | API key and model. |
| Custom Provider | At the OpenAI-compatible endpoint you enter | Optional | Server URL and Model ID. |
The model choices shown by RenamerX can change as Provider catalogs change. Choose a listed compatible model or enter a custom Model ID when the configuration allows it.
For an editable server URL, use a complete HTTP or HTTPS URL without embedded credentials, a query, or a fragment. Use HTTPS for a remote endpoint. Ollama, LM Studio, and Custom Provider are only local when the configured endpoint is local.
Understand the data boundary
RenamerX sends requests directly from the desktop app to the active Provider. It does not relay connected Provider requests through a RenamerX cloud proxy.
Depending on the file and template, a request can include:
- extracted document text
- normalized file metadata, which can include the original filename and local path
- the naming fields, output language, and custom instructions from the selected template
- a visual input prepared from a supported image, scanned PDF, or video
- a human-readable place name from file metadata or offline GPS resolution, when available
RenamerX removes raw GPS coordinates from the metadata prepared for AI processing. A readable location can still be included. OpenAI and Google Gemini process the request in their cloud services. Ollama, LM Studio, and Custom Provider process it at the server URL you configured.
Connect and activate a Provider
Open Preferences → AI Provider & Model, then complete the fields shown for your AI Provider.
- Select an AI Provider.
- Select a model or enter a Model ID when required.
- Check the Server URL when it is editable.
- Enter an API Key when required by the Provider or endpoint.
- Choose Test connection if you want to validate the current fields without activating them.
- Choose Save & Activate to test the configuration and make it active.
Save & Activate always runs the connection test before activation. If the test fails, RenamerX does not activate the new configuration. When you select Built-in Local AI and its managed resources are missing, RenamerX guides you through the required download before activation can finish.
What the connection test verifies
The connection test sends a real image request using the selected endpoint, credentials, and model. It succeeds when the request completes within the test timeout and returns non-empty model text.
A successful test confirms only that this configuration completed the request. It does not require a fixed JSON response or evaluate suggested filename quality. It also does not guarantee support for every file or future service availability. If a configured service later becomes unreachable, test it again from Preferences.
Switch Providers without mixing results
RenamerX reads the active AI Provider when each file reaches the AI processing step. It does not lock one AI Provider to an entire batch or processing queue.
Files already waiting or being analyzed can use different AI Providers if you switch during processing. Stop or finish current processing before switching when a group of files must use one AI Provider. Suggested filenames that are already ready do not change automatically. Retry those files if you want new results from the new AI Provider.
Protect keys and account usage
RenamerX stores AI Provider API keys in a dedicated encrypted file on the device. Keys are not stored in normal preferences, logs, diagnostic bundles, or cached naming results.
Removing a key takes effect immediately. Enter the replacement key and use Save & Activate again if the AI Provider requires authentication.
Third-party quotas, billing, privacy terms, and service availability remain the AI Provider's responsibility. A RenamerX Pro license does not include OpenAI, Google Gemini, or other third-party usage.