Skip to main content

Chat Setup

Prerequisites​

  1. Voyage AI API key — for embedding queries and reranking
  2. MongoDB Atlas — with embedded documents (use vai pipeline to ingest)
  3. LLM provider — Anthropic, OpenAI, or Ollama

Configure Your LLM​

Anthropic (Claude)​

vai config set llm-provider anthropic
vai config set llm-api-key sk-ant-your-key-here
vai config set llm-model claude-sonnet-4-20250514

OpenAI (GPT)​

vai config set llm-provider openai
vai config set llm-api-key sk-your-key-here
vai config set llm-model gpt-4

Ollama (Local)​

# Install Ollama first: https://ollama.ai
ollama pull llama3

vai config set llm-provider ollama
vai config set llm-base-url http://localhost:11434
vai config set llm-model llama3

Start Chatting​

vai chat --db myapp --collection docs

Using CLI Flags (No Config)​

vai chat --db myapp --collection docs \
  --llm-provider anthropic \
  --llm-api-key sk-ant-your-key \
  --llm-model claude-sonnet-4-20250514

Project Config​

You can also set chat options in .vai.json:

{
  "db": "myapp",
  "collection": "docs",
  "chat": {
    "mode": "pipeline",
    "maxContextDocs": 5,
    "maxConversationTurns": 20
  }
}

Verify Setup​

# Check all config
vai config list

# Test connectivity
vai ping

# Estimate per-turn cost
vai chat --estimate