Client Setup

TokOne is compatible with the OpenAI, Anthropic, and Gemini protocols, so any client that lets you set a custom Base URL + API Key can connect. Two steps:

1. Point the client's Base URL at TokOne (e.g. https://api.tokone.ai/v1).
2. Set the API Key to your sk- key.

Below is the general approach by client type.

General configuration

Setting	Value
Base URL / API address	https://api.tokone.ai/v1 (OpenAI, Anthropic compatible) https://api.tokone.ai/v1beta (native Gemini)
API Key	your sk-... key
Model	see Models

TIP

Some clients auto-append /v1, others don't. If you get a 404, check that the final path is .../v1/chat/completions — not missing or duplicating /v1.

CLI tools

For command-line tools configured via environment variables. Common form:

export OPENAI_API_KEY="sk-..."
export OPENAI_BASE_URL="https://api.tokone.ai/v1"

For Anthropic-style CLIs:

export ANTHROPIC_API_KEY="sk-..."
export ANTHROPIC_BASE_URL="https://api.tokone.ai"

Editors / IDEs

Most AI coding plugins (the ones with an "OpenAI compatible" or "custom provider" option) have:

• API Base / Endpoint: https://api.tokone.ai/v1
• API Key: sk-...
• Model: pick one from Models

Save and you're ready to use completion / chat.

Desktop chat clients

For desktop clients that support a custom OpenAI endpoint:

1. Create a new "OpenAI compatible" provider.
2. Set endpoint to https://api.tokone.ai/v1.
3. Set key to sk-....
4. Fetch or type in a model name, then start chatting.

Don't see your client?

As long as the client lets you change the Base URL, it can use TokOne. To pick the protocol:

• Client says it's OpenAI compatible → use /v1, header Authorization: Bearer.
• Client is Claude / Anthropic style → use /v1 (/v1/messages), header x-api-key.
• Client is Gemini style → use /v1beta, header x-goog-api-key.

Still stuck? See API Endpoints or the FAQ.