VoiceVision - AI Assistant with custom chat API

How to Use Custom Chat API in Your AI Assistant with VoiceVision

In today’s AI-driven world, businesses need flexible and customizable AI assistants to engage users effectively. VoiceVision provides a powerful solution by allowing you to integrate custom chat APIs into your AI assistant. Whether you’re using Perplexity, DeepSeek, Anthropic, or your own custom LLM chat API, this feature ensures that your AI assistant responds intelligently and accurately.

In this tutorial, we’ll walk you through the step-by-step process of integrating a custom chat API into your VoiceVision assistant.

What Is a Custom Chat API in VoiceVision?

A custom chat API allows your AI assistant to use an external AI model for generating responses instead of relying on built-in models. This feature provides flexibility by enabling you to:

✅ Connect third-party AI models such as Perplexity, DeepSeek, Anthropic, or any other LLM API.

✅ Use a proprietary AI model designed for your specific business needs.

✅ Have full control over response generation, including tone, accuracy, and personality.

🚨 Note: Custom chat APIs are not supported in RAG-based assistants. Ensure that your assistant is not using a retrieval-augmented generation (RAG) model before enabling this feature.

Step 1: Accessing Assistant Settings

1️⃣ Log in to your VoiceVision dashboard.

2️⃣ Navigate to the Assistant Page.

3️⃣ Select the assistant you want to modify.

4️⃣ Click on the Settings icon.

Step 2: Enabling the Custom Chat API

1️⃣ Go to the Advanced Settings tab.

2️⃣ Scroll down to the Custom Chat Completion API Settings section.

3️⃣ Toggle the switch to Enable Custom Chat Completion API.

Step 3: Configuring API Details

Once you’ve enabled the feature, enter the following details:

🔹 API URL – The endpoint where chat requests will be sent. This is the main connection between your AI assistant and the custom AI model.

🔹 API Key – This key is required to authenticate and securely access the API.

🔹 AI Model – Specify the model name that will process the chat completion request (e.g., GPT-4, Perplexity-XL, DeepSeek, etc.).

📌 Example Configuration:

API URL: https://your-custom-api.com/chat
API Key: sk-abc123xyz
AI Model: Perplexity-XL

4️⃣ Click Save to apply the changes.

Step 4: Testing Your AI Assistant

Now that your assistant is connected to a custom chat API, it’s time to test it in action:

1️⃣ Go to the Assistant Chat Page.

2️⃣ Start a conversation using text or voice.

3️⃣ Observe the responses and check if the API integration is working correctly.

💡 If your assistant is not responding as expected, double-check your API URL, API Key, and Model Name in the settings.