How are tokens billed?
Input and output tokens are metered per model using each model’s native tokenizer. Rates depend on your plan and the model price list.
Pricing
Transparent token pricing. Upgrade for higher limits, priority support, and volume discounts.
| Plan | Input ($/1M) | Output ($/1M) | Highlights |
|---|---|---|---|
| Basic | $0.20 | $0.40 | 1M free tokens · standard support · baseline uptime |
| Pro | $0.15 | $0.30 | Higher limits · priority support · volume discount when monthly spend qualifies · 99.9% uptime target |
| Enterprise | Custom | Custom | Dedicated clusters · custom models · SLA · dedicated support |
FAQ
Do you charge for failed requests?
Failed requests that never return a completion are not billed. Successful completions after routing retries are billed once.
How do I use my OpenAI SDK?
Set base_url to the TokenHub API endpoint and use model IDs from our catalog. See Quickstart.
Data & privacy
See Privacy and DPA. Enterprise customers can discuss residency and processing terms with our team.