AiChat CLI

A command-line tool for AI Dialogue via the AceDataCloud platform.

Features

• Chat with AI — Send questions to GPT, DeepSeek, Grok, GLM, and more models
• Conversation Continuation — Resume conversations using --id
• Stateful Sessions — Enable multi-turn context with --stateful
• References — Inject context documents into prompts with --ref
• Rich Output — Beautiful terminal formatting with --json for scripting
• Many Models — GPT-5, GPT-4o, o3, o4-mini, DeepSeek, Grok, GLM, and more

Installation

pip install aichat-pro-cli

Quick Start

# Set your API token
export ACEDATACLOUD_API_TOKEN=your_token_here

# Ask a question
aichat chat "What is the capital of France?"

# Use a specific model
aichat chat "Explain quantum computing" -m gpt-4o

# Continue a conversation
aichat chat "Tell me more" --id 64a67fff-61dc-4801-8339-2c69334c61d6

# Stateful conversation (server remembers context)
aichat chat "My name is Alice" --stateful
aichat chat "What is my name?" --id <returned-id> --stateful

# With reference documents
aichat chat "Summarize this" --ref "https://example.com/doc.txt"

# Get JSON output
aichat chat "Hello" --json | jq '.answer'

Commands

Chat Options

Chat2 Turn Limit

aichat chat2 --max-turns accepts integers from 1 to 500 and limits agent iterations (model calls), not conversation history. When omitted, the CLI leaves max_turns out of the request and the server defaults to 500. Each model call is billed by actual usage; a higher limit allows longer tool chains.

Set the limit to 1 for a single answer without tool calls:

aichat chat2 "Explain quantum computing" --max-turns 1

Configuration

You can also use a .env file or pass --token directly.

Docker

docker compose run aichat-cli chat "Hello!"

Development

# Install with dev dependencies
pip install -e ".[all]"

# Run tests
pytest

# Run linter
ruff check .
ruff format --check .

License

MIT License - see LICENSE for details.

Links

• AceDataCloud Platform
• API Documentation