diff --git a/README.md b/README.md index e9f3b3cf..8c1a26f4 100644 --- a/README.md +++ b/README.md @@ -54,7 +54,7 @@ Check out the demo video below for the first sneak peak of what Copilot Plus can - Take advantage of powerful **User Custom Prompts**! You can _add, apply, edit, and delete_ your custom prompts, all stored locally in your Obsidian vault as markdown. Trigger any custom prompt with a simple `/` in the chat input! - **Interact with your entire vault** using "Vault QA" mode. Ask questions and receive **cited responses**! - All QA modes leverage retrieval augmentation with a **local index**, ensuring your data isn't sent to a cloud-based vector search service. -- **Copilot Plus (in limited Alpha)**: supercharge your second brain with a **powerful AI agent** right inside your Obsidian vault. More powerful than anything else out there. [Learn more here](https://obsidiancopilot.com). +- **Copilot Plus (beta)**: supercharge your second brain with a **powerful AI agent** right inside your Obsidian vault. [Learn more here](https://obsidiancopilot.com). Chat and Vault QA modes are completely free to use. Copilot Plus is currently in limited alpha and free to use for alpha testers. @@ -66,19 +66,15 @@ Chat and Vault QA modes are completely free to use. Copilot Plus is currently in ## 🤗 New to Copilot? Quick Guide for Beginners -The simplest way to get started is to get an OpenAI API key (with payment method set up with OpenAI) and set it in the Copilot settings. Then simply pick `gpt-4o` or `gpt-4o-mini` in the model dropdown to start chatting. Please check out the [documentation](https://obsidiancopilot.com/en/docs) for more details. +The simplest way to get started is to get an OpenAI API key (with payment method set up with OpenAI) and set it in the Copilot settings. Then simply pick `gpt-4o` or `gpt-4o-mini` in the model dropdown to start chatting. Make sure to set your **default chat and embedding model** in the first tab of Copilot settings. + +For plus users, you can keep the embedding model as is `copilot-plus-small` (it's a smaller but better model than the OpenAI large, and it's covered by your subscription). For free users, you can set the embedding model to `text-embedding-3-small` as the starting point, and explore other embedding models once you get a sense of how this works. + +Please check out the [documentation](https://obsidiancopilot.com/en/docs) for more details. Try going through all the Copilot commands in the command palette, and see what they do. You can also checkout my [youtube channel](https://youtube.com/@loganhallucinates?sub_confirmation=1) for previous demos. -#### 🔧 Copilot Settings - -Explore the "General Settings" to select your default chat model and customize parameters such as temperature, max tokens, and conversation context to fit your needs. - -Settings - -Enter your API keys in "API settings" for existing providers. - -API Settings +#### 🔧 Adding Models Add model using the "Custom Model" form. diff --git a/images/api-settings.png b/images/api-settings.png deleted file mode 100644 index ed879916..00000000 Binary files a/images/api-settings.png and /dev/null differ diff --git a/images/settings.png b/images/settings.png deleted file mode 100644 index 08d13dc4..00000000 Binary files a/images/settings.png and /dev/null differ