travisvn_obsidian-vision-recall/docs/reference.md

API Key Setup for Screenshot AI (Obsidian Plugin)

This guide will help you obtain API keys from supported providers and configure them for use with the Screenshot AI Obsidian plugin.

1. Choosing Your AI Models

To process screenshots effectively, the plugin requires:

• A vision model: Used to extract and analyze image content.
• A regular LLM model: Used for text-based queries and reasoning. This model may also support vision (i.e., multimodal models like GPT-4 Turbo with Vision).

You can use the same model for both tasks or specify separate models.

2. Obtaining an API Key

OpenAI

1. Visit OpenAI's API page.
2. Sign in or create an account.
3. Go to API Keys and generate a new key.
4. Copy your API key and store it securely.

OpenRouter (Alternative API Gateway)

OpenRouter provides a unified API to access multiple models from different providers.

1. Go to OpenRouter's API page.
2. Sign in or create an account.
3. Navigate to the API keys section and generate a key.
4. Copy the API key and store it securely.

3. Configuring the Plugin

Once you have obtained an API key, you need to configure it within the plugin settings.

Setting Up API Keys

In Obsidian:

1. Open Settings.
2. Navigate to Screenshot AI.
3. Locate the API Key field and enter your key.

Specifying Models

You need to define:

• Vision Model: A model capable of processing images (e.g., gpt-4-turbo, gpt-4-vision-preview, openrouter/llava).
• LLM Model: A text model for answering queries (e.g., gpt-4-turbo, mistral, claude-3).

Example Configuration:

{
  "api_key": "your-api-key-here",
  "vision_model": "gpt-4-vision-preview",
  "llm_model": "gpt-4-turbo"
}

4. Notes on API Usage & Limits

• OpenAI and OpenRouter have different pricing structures. Check their respective pricing pages.
• OpenRouter allows routing requests to different providers.
• Ensure your API key has sufficient quota to avoid service interruptions.

For troubleshooting, refer to the Obsidian Console (Ctrl+Shift+I on Windows/Linux, Cmd+Option+I on macOS) to check for API-related errors.

---

For further assistance, visit our GitHub Issues page or the plugin documentation.