codex/sdk/python/docs/getting-started.md

Getting Started

This is the fastest path from install to a multi-turn thread using the public SDK surface.

The SDK is experimental, so the public API and runtime requirements may keep evolving before the first public release.

1) Install

From repo root:

cd sdk/python
uv sync
source .venv/bin/activate

Requirements:

• Python >=3.10
• uv
• installed openai-codex-cli-bin runtime package, or an explicit codex_bin override

2) Authenticate when needed

Existing Codex auth state is reused automatically. To authenticate from the SDK, use the flow that fits your app:

from openai_codex import Codex

with Codex() as codex:
    codex.login_api_key("sk-...")
    account = codex.account()
    print(account.account)

Interactive ChatGPT browser login returns a handle that carries the URL and the matching completion event:

with Codex() as codex:
    login = codex.login_chatgpt()
    print(login.auth_url)
    completed = login.wait()
    print(completed.success)

Device-code login works the same way with login_chatgpt_device_code(), which exposes verification_url, user_code, and wait().

3) Run your first turn (sync)

from openai_codex import Codex

with Codex() as codex:
    server = codex.metadata.serverInfo
    print("Server:", None if server is None else server.name, None if server is None else server.version)

    thread = codex.thread_start(model="gpt-5.4", config={"model_reasoning_effort": "high"})
    result = thread.run("Say hello in one sentence.")

    print("Thread:", thread.id)
    print("Text:", result.final_response)
    print("Items:", len(result.items))

What happened:

• Codex() started and initialized codex app-server.
• thread_start(...) created a thread.
• thread.run("...") started a turn, consumed events until completion, and returned TurnResult with turn metadata, final assistant response, collected items, and usage.
• result.final_response is None when no final-answer or phase-less assistant message item completes for the turn.
• use thread.turn(...) when you need a TurnHandle for streaming, steering, or interrupting before collecting TurnResult
• one client can consume multiple active turns concurrently; turn streams are routed by turn ID

4) Continue the same thread (multi-turn)

from openai_codex import Codex

with Codex() as codex:
    thread = codex.thread_start(model="gpt-5.4", config={"model_reasoning_effort": "high"})

    first = thread.run("Summarize Rust ownership in 2 bullets.")
    second = thread.run("Now explain it to a Python developer.")

    print("first:", first.final_response)
    print("second:", second.final_response)

5) Async parity

Use async with AsyncCodex() as the normal async entrypoint. AsyncCodex initializes lazily, and context entry makes startup/shutdown explicit.

import asyncio
from openai_codex import AsyncCodex


async def main() -> None:
    async with AsyncCodex() as codex:
        thread = await codex.thread_start(model="gpt-5.4", config={"model_reasoning_effort": "high"})
        result = await thread.run("Continue where we left off.")
        print(result.final_response)


asyncio.run(main())

6) Resume an existing thread

from openai_codex import Codex

THREAD_ID = "thr_123"  # replace with a real id

with Codex() as codex:
    thread = codex.thread_resume(THREAD_ID)
    result = thread.run("Continue where we left off.")
    print(result.final_response)

7) Public app-server types

The convenience wrappers live at the package root. Public app-server value and event types live under:

from openai_codex.types import ThreadReadResponse, Turn, TurnStatus

8) Next stops

• API surface and signatures: docs/api-reference.md
• Common decisions/pitfalls: docs/faq.md
• End-to-end runnable examples: examples/README.md