Frequently asked questions
Everything about models, pricing, billing and the API.
What is Tokion?
Tokion is a single API for every major AI model: GPT, Claude, Gemini, DeepSeek, Kimi, Grok and more, plus image and video generation. One key, one balance, one OpenAI-compatible endpoint.
How is Tokion cheaper than official APIs?
We buy capacity directly from multiple suppliers and sell it by usage — no subscriptions, no minimums, no credit markups. On some models the savings reach 210× versus official pricing.
Is it really OpenAI-compatible?
Yes. Point your existing OpenAI SDK at https://tokion.io/api/v1, replace the key, and your code keeps working. Anthropic-format /v1/messages is supported too, so Claude Code, Cursor and Cline work out of the box.
Can I try it for free?
Yes — new accounts get free tokens for their first requests, and the web chat works without signup. No credit card required.
How does billing work?
You top up a balance and each request deducts only the tokens it used. Failed upstream calls are not billed. Full usage history is available in the dashboard.