Skip to content

Guide to Using Obsidian Copilot via ZenMux

Obsidian Copilot is a powerful AI assistant plugin. Through its integration with ZenMux, you can use multiple advanced AI models within your Obsidian notes for a smarter writing and thinking experience.

Compatibility Note

ZenMux fully supports the OpenAI API protocol and can be seamlessly integrated into the Obsidian Copilot plugin. Minimal configuration is required.

Note that the OpenAI protocol base_url="https://zenmux.ai/api/v1".

Installation and Configuration

1. Install Obsidian

Go to the Obsidian website to download and install the Obsidian desktop app.

System Support

Obsidian supports Windows, macOS, and Linux. Choose the version appropriate for your system.

2. Install the Copilot Plugin

  1. Open the Obsidian app
  2. Click the Settings icon (gear icon) in the bottom-left corner
  3. In the Settings page, find Community plugins
  4. Click the Browse button to open the plugin community

Browse community pluginsSearch community plugins

  1. Enter copilot in the search box
  2. Find the Obsidian Copilot plugin and click Install

Install Copilot plugin

  1. After installation, click Enable to activate the plugin

3. Configure ZenMux Models

Get Your ZenMux API Key

Important Configuration

Before configuration, you need to obtain a ZenMux API Key. Visit the ZenMux Console to create or view your API Key.

Add a Custom Model

  1. Open the Obsidian Settings page
  2. In the left menu, find Community pluginsCopilot plugin settings
  3. Click the Models tab
  4. Click the Add model button

Add model button

Fill in the Model Configuration

In the pop-up configuration form, fill in the fields as follows:

Model list page

  • Provider: Select OpenAI
  • Base URL: Enter https://zenmux.ai/api/v1
  • API Key: Enter your ZenMux API Key (starts with sk-ai-v1-)
  • Model ID: Enter the model slug on ZenMux that supports the OpenAI Chat-Completion protocol

4. Test the Model Connection

After configuration, click the Test button to verify connectivity:

If configured correctly, you will see a success message. If the connection fails, see the Troubleshooting section.

Using Copilot

Launch the Chat Interface

After configuration, you can use Copilot in the following way:

  1. Click the Copilot icon (robot icon) in the left sidebar
  2. Open the chat interface
  3. Select your configured ZenMux model in the model selector

Use Copilot chatSelect model

Troubleshooting

Common Issue Resolution

API Key Error

Issue: API Key is invalid or authentication failed

Solution:

  • Check whether the API Key was copied correctly and ensure there are no extra spaces
  • Verify the API Key format starts with sk-ai-v1-
  • Confirm the API Key is active and has sufficient balance
  • Check the API Key status in the ZenMux Console
Connection Failure

Issue: Copilot cannot connect to the ZenMux service

Solution:

  • Check if your network connection is working properly
  • Verify the Base URL is configured as https://zenmux.ai/api/v1 (note it is /v1 and not /anthropic)
  • Confirm your firewall or proxy settings are not blocking the connection
  • Try accessing https://zenmux.ai in a browser to confirm the service is reachable
Model Unavailable

Issue: The model is reported as non-existent or does not support the OpenAI protocol

Solution:

  • Check that the Model ID is correct, paying attention to case and spelling
  • Common mistake: anthropic/claude-hiku-4.5-sonnet should be anthropic/claude-haiku-4.5 (haiku, not hiku)
  • Visit the ZenMux Model List and filter by "OpenAI API Compatible" to view available models
  • Confirm the selected model supports the OpenAI Chat-Completion protocol
Test Succeeds but Fails During Use

Issue: Connection test succeeds, but errors occur during actual use

Solution:

  • Check that your API Key has sufficient balance
  • Verify the model capability settings are correct (Chat, Streaming, etc.)
  • Open the Obsidian Developer Console (Ctrl/Cmd + Shift + I) for detailed error messages
  • Try switching to another model to test whether it is a model-specific issue
Slow Response or Timeout

Issue: AI responses take too long or requests time out

Solution:

  • Check your network connection quality
  • Try faster models such as anthropic/claude-haiku-4.5 or deepseek/deepseek-chat
  • Reduce the text length per conversation
  • Adjust timeout parameters in Copilot settings
Plugin Configuration Lost

Issue: Configuration is lost after restarting Obsidian

Solution:

  • Ensure the Copilot plugin is properly enabled
  • Check permissions for the .obsidian/plugins/obsidian-copilot/ directory in your Obsidian vault
  • Try reinstalling the Copilot plugin
  • Back up important configurations to avoid accidental loss

Contact Us

If you encounter any issues during use, or have suggestions and feedback, feel free to contact us:

For more contact methods and details, please visit our Contact Us page.