lux ai

lux ai

The ai command provides tools for interacting with the Lux AI network, including chat completion, model listing, and agent execution.

Connects to a local lux-ai node or the Hanzo AI gateway (api.hanzo.ai).

ENDPOINTS:

Local: http://localhost:9090 (default, via lux ai serve) Gateway: https://api.hanzo.ai/v1

ENVIRONMENT:

LUX_AI_ENDPOINT Override the default AI endpoint LUX_AI_API_KEY API key for gateway authentication

Usage:

lux ai

lux ai agent

Run an AI agent that executes a task using the connected AI endpoint.

The agent sends the task as a system-prompted chat request with tool-use capabilities when supported by the model.

Examples: lux ai agent "Deploy a new EVM chain on devnet" lux ai agent --model qwen3-8b "Analyze the validator set"

Usage:

lux ai agent [task] [flags]

Flags:

      --model string   Model to use (default "qwen3-8b")

lux ai chat

Send a chat message to the AI model and print the response.

Examples: lux ai chat "What is the Lux network?" lux ai chat --model qwen3-8b "Explain post-quantum cryptography" lux ai chat --system "You are a blockchain expert" "What is BFT?"

Usage:

lux ai chat [message] [flags]

Flags:

      --model string    Model to use (default "qwen3-8b")
      --system string   System prompt

lux ai complete

Generate a text completion from a prompt.

Examples: lux ai complete "The Lux blockchain uses" lux ai complete --model zen-coder-1.5b --max-tokens 256 "func main() {"

Usage:

lux ai complete [prompt] [flags]

Flags:

      --max-tokens int   Maximum tokens to generate (0 = model default)
      --model string     Model to use (default "qwen3-8b")

lux ai models

List all models available on the connected AI endpoint.

Examples: lux ai models LUX_AI_ENDPOINT=https://api.hanzo.ai lux ai models

Usage:

lux ai models