Docs.
Detailed guides for every way to connect to 2KEN. Use the same token, base URL, and model configuration from command lines, OpenAI-compatible SDKs, HTTP APIs, code clients, and switching tools.
Choose an integration path
Select the tool or runtime you use from the sidebar.
Basic calls
Call 2KEN directly with cURL from Bash on macOS, Linux, or WSL. Use this to verify the token, base URL, and model name.
PowershellSend an OpenAI-compatible request from Windows PowerShell to confirm network access, authentication, and model output.
CMDUse cURL from Windows CMD to call 2KEN when PowerShell scripting is not available.
OpenAI SDKConnect any OpenAI-compatible SDK to the 2KEN API address. Replace only `baseURL` and the token while keeping chat completions familiar.
HTTP APISend requests directly to the 2KEN HTTP API. This path fits backend services, gateways, low-code platforms, and custom clients.
Clients
Configure the 2KEN-compatible address, token, and default model for Claude Code from a Windows terminal.
macOSExport the Claude Code variables in a macOS terminal session and route requests through the 2KEN-compatible endpoint.
LinuxWrite the Claude Code-compatible address and token into a Linux shell profile for remote development machines or containers.
WindowsConfigure Codex on Windows with an OpenAI-compatible base URL, API key, and default model.
macOSUse OpenAI-compatible environment variables on macOS so Codex routes model requests through 2KEN.
LinuxSave the 2KEN OpenAI-compatible configuration in a separate env file on Linux, then load it into Codex sessions as needed.
WindowsPrepare an OpenAI-compatible provider profile for OpenClaw on Windows and use 2KEN as the unified model entry point.
macOSRoute OpenClaw requests through 2KEN on macOS with an OpenAI-compatible provider profile.
LinuxUse environment variables or the client provider settings on Linux so OpenClaw uses the 2KEN unified endpoint.
Cherry StudioAdd an OpenAI Compatible provider in Cherry Studio, then enter the 2KEN API address, token, and model ID.
Admin tools
Ready to ship?
Open the console next, copy a token, inspect model policy, and place the real values into the matching client.