License key: The LrGeniusTagAI license key you received when buying the plugin.

AI Model: The AI model/engine you want to use. In order to make Ollama models show up here, you have to install Ollama and pull a model.

Show costs: When using ChatGPT or Gemini models, costs are generated based on your usage of the AI. If this checkbox is activated, after each AI run a cost summary is shown. I won't guarantee that it is 100% accurate.

API keys / Ollama base URL:

• Create Google API key via "Get API key" button and enter value in the text field to the left. (Only if you want to use Google Gemini as backend)

• Create ChatGPT API key via "Get API key" button and enter value in the text field to the left. (Only if you want to use ChatGPT as backend)

• Ollama base URL: Leave default value in most cases. (READ)
  If you have not setup Ollama yet hit the button or click here: Ollama setup (Only if you want to use Ollama as backend)

Prompts: In this section it is possible to adapt the AI prompt to your needs.

Result language: Language of the generated contents. For now English, French and German are supported.

Generate the following: Decide which content should be generated. Reducing the amount of generated content, will reduce costs and AI runtime duration.

Validate before saving: Each generated text can be edited before it is saved to the catalog. This isn't implemented for keywords.

Submit existing metadata: Existing GPS and keywords can be handed to the AI. GPS coordinates help the AI recognize landmarks for example.

Show preflight dialog: Before each AI run a dialog is displayed allowing to verify / change the used settings for the AI. So you don't need to go to the Plug-In manager every time.

Show photo context dialog: Before the analysis of each image you have the chance to enter some additional information for the AI. This is also possible in the metadata panel of Lightroom.

Keyword hierarchy: Put generated keywords beneath keyword categories.

Edit keyword categories: Change the categories names to your needs.

Use top-level keyword: If activated all generated keywords are put hierarchically beneath a keyword called: ChatGPT, Google Gemini, Ollama. According to the used AI. This makes it easier to distinguish generated keywords from manually created ones. Also managing the keywords is easier in the "Keyword list" field from Lightroom classic.

Export settings: The plugin doesn't send your RAW images to the AI. Instead a reduced size export is used. Here you can modify the settings for that, to enhance quality or reduce bandwith usage.