Developer docs

Connect Claude Code, Codex, and Gemini CLI to Modelane

Modelane exposes a unified inference endpoint at https://api.modelane.ai/v1. Use one key, one balance, and one routing console across frontier model workflows.

TL;DR

  • Use https://api.modelane.ai/v1 for OpenAI-compatible tools.
  • Run an installer script to write local CLI configuration.
  • Create separate keys per tool so each key can be rotated safely.

Claude-native

Claude Code setup

Configure Claude Code-style workflows to use Modelane as the Claude-compatible API endpoint.

Open guide

OpenAI-compatible

Codex CLI setup

Use Modelane with OpenAI-compatible coding tools by setting one API key and base URL.

Open guide

Gemini-compatible

Gemini CLI setup

Configure Gemini-oriented workflows with Modelane routing and a stable key management path.

Open guide

OpenAI-compatible quickstart

This is the shortest path for SDKs and CLIs that support a custom OpenAI base URL.

export MODELANE_API_KEY="ml_..."

python - <<'PY'
import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["MODELANE_API_KEY"],
    base_url="https://api.modelane.ai/v1",
)

response = client.chat.completions.create(
    model="modelane-fast",
    messages=[{"role": "user", "content": "Hello, Modelane."}],
)
print(response.choices[0].message.content)
PY