cli2api 0.2.0

OpenAI-compatible API over Claude Code CLI

CLI2API

OpenAI-compatible API over Claude Code CLI.

Why CLI2API?

Claude Code CLI is powerful but not all tools support it directly. CLI2API bridges this gap by exposing Claude Code as an OpenAI-compatible API.

Use Cases:

• IDE Integration — Connect Kilo Code, Roo Code, Cursor, or other editors that support OpenAI API
• Custom Applications — Build apps using familiar OpenAI SDK instead of spawning CLI processes
• Team Sharing — Run one CLI2API server and share Claude access across your team
• Tool Compatibility — Use Claude with any tool that supports OpenAI API format (LangChain, AutoGPT, etc.)

Requirements

• Python 3.11+
• Claude Code CLI installed and authenticated

Installation

Quick Start (pip)

pip install cli2api
cli2api

Server starts at http://localhost:8000

From Source

git clone https://github.com/anoxis/CLI2API.git
cd CLI2API
pip install -e .
cli2api

Docker

Docker requires the repository to be cloned first:

# Clone and build
git clone https://github.com/anoxis/CLI2API.git
cd CLI2API
docker build -t cli2api .

# Run with Claude CLI mounted from host
docker run -p 8000:8000 \
  -v $(which claude):/usr/local/bin/claude:ro \
  -e CLI2API_CLAUDE_CLI_PATH=/usr/local/bin/claude \
  cli2api

Or use docker-compose:

# Edit docker-compose.yaml to mount your Claude CLI path
docker-compose up -d

Development

pip install -e ".[dev]"
pytest tests/ -v

IDE Integration

Kilo Code / Roo Code

1. Start CLI2API server: cli2api
2. Open extension settings
3. Add custom provider:
   • Provider Name: CLI2API (or any name)
   • API Base URL: http://localhost:8000/v1
   • API Key: not-needed (any non-empty value)
   • Model: sonnet (or opus, haiku)

Cursor / Continue

1. Start CLI2API server
2. Settings → Models → Add Model
3. Configure:
   • API Base: http://localhost:8000/v1
   • API Key: any-value
   • Model: sonnet

Generic OpenAI-compatible client

from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:8000/v1",
    api_key="not-needed"
)

response = client.chat.completions.create(
    model="sonnet",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)

Usage

Test with curl

# Health check
curl http://localhost:8000/health

# List models
curl http://localhost:8000/v1/models

# Chat completion
curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model": "sonnet", "messages": [{"role": "user", "content": "Hello"}]}'

# Streaming
curl http://localhost:8000/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model": "sonnet", "messages": [{"role": "user", "content": "Hello"}], "stream": true}'

Configuration

Environment variables (prefix CLI2API_):

Variable	Default	Description
HOST	0.0.0.0	Server host
PORT	8000	Server port
CLAUDE_CLI_PATH	auto-detect	Path to Claude CLI executable
DEFAULT_TIMEOUT	300	CLI execution timeout (seconds)
DEFAULT_MODEL	sonnet	Default model
CLAUDE_MODELS	sonnet,opus,haiku	Available models (comma-separated)
LOG_LEVEL	INFO	Logging level
LOG_JSON	false	JSON format logging
DEBUG	false	Debug mode

Copy .env.example to .env for local configuration.

Available Models

Model ID	Description
sonnet	Claude Sonnet (default)
opus	Claude Opus
haiku	Claude Haiku

API Endpoints

Endpoint	Description
GET /health	Health check
GET /	API info
GET /v1/models	List available models
GET /v1/models/{id}	Get model info
POST /v1/chat/completions	Chat completions (OpenAI compatible)
POST /v1/responses	Responses API (OpenAI compatible)

Security Notice

CLI2API does not implement authentication. It is designed for local use only.

• Do not expose to the internet without additional security measures
• Use behind a reverse proxy with authentication if network access is needed
• The API inherits permissions from the Claude CLI authentication on the host

License

MIT