The Mistral AI block enables seamless interaction with Mistral’s AI models, allowing users to create chat completions based on their queries and display responses in the Agent Dialog.

​
Creating a Mistral Account

To get started with Mistral AI:

  1. Visit Mistral Console.
  2. Create an account and log in.

​
Chat Completion with Mistral AI

Using the Mistral AI block, you can generate chat completions based on user queries. The integration supports the Dialogue message type, which allows the assistant to maintain a conversation history for a smooth and contextualized interaction.

​
Steps to Use Mistral AI for Chat Completion

  1. Set Up the Integration:

    • Access the Mistral AI block within your project.
    • Configure the block by adding your Mistral API key.

  2. Prepare User Inputs:

    • Collect user messages and organize them in a sequence.
    • Ensure a proper format for passing the conversation history.

  3. Generate Responses:

    • Send the prepared dialogue sequence to the Mistral API.
    • Display the response in the Agent Dialog.

  4. Test the Integration:

    • Verify that user queries receive appropriate and timely responses.

​
Example Configuration

Below is an example JSON payload to interact with Mistral’s chat completion API:

{
  "messages": [
    { "role": "user", "content": "Hello, how can you help me?" },
    { "role": "assistant", "content": "I can assist with various tasks. What do you need help with?" }
  ],
  "model": "mistral-7b-v0.1",
  "temperature": 0.7
}

​
Notes on Dialogue Message Type

  • Dialogue allows passing a sequence of saved messages between the user and assistant.
  • Ensure the conversation history is structured in a chronological format.

​
Troubleshooting Common Issues

​
HTTP 401 Error

If you encounter an HTTP 401 error when loading Mistral models:

  • Your API key might not be fully propagated.
  • Wait a few minutes and try again.

​
Rate Limit Errors

If you see a rate limit error:

  • Check your subscription status.
  • Visit the Mistral Billing Section and click the Subscribe button to upgrade your plan.

​
Validation and Testing

Before deploying the Mistral AI integration:

  • Verify the API key is correctly set.
  • Ensure the chat completions return accurate and relevant responses.
  • Test the integration under various scenarios to validate performance.

Note: Regularly monitor your usage to avoid exceeding rate limits or subscription restrictions.