Model fees are fixed by model
The same model has the same base fee across plans. Plans do not change model unit price.
Transparent model pricing, agent-friendly strategy, global service, and enterprise-grade delivery.
Simple, consistent, and designed for long-term production growth.
The same model has the same base fee across plans. Plans do not change model unit price.
Total cost = model invocation fee + 3% platform service fee.
Top-up bonuses, referral rewards, and TokenPlan subscription discounts are available.
Supports capability-priority and price-priority routing, plus cache-first strategy with cache pricing.
Pro supports service-fee waiver, and large-scale usage can unlock additional discounts.
Billing is settled in USD and supports multiple recharge methods.
If you need a detailed walkthrough of plan differences, please Contact us, and a specialist will follow up.
You can choose the United States, the European Union, or Global. If the selected region has no supplier for a model, invocation may fail.
When a region lacks direct model supply, enterprise teams can use data desensitization workflows, equivalent-model substitution consulting, or dedicated delivery plans.
For strategic partnerships, global compute-center deployment can be provided.
USD is the core billing currency for both recharge and consumption.
Each successful invocation is billed by model usage, then the service fee policy is applied based on your plan.
The default is a 3% platform service fee. Pro and Enterprise may apply fee waivers or custom terms under agreement.
If there is no supplier in that region for the selected model, the call may fail. Enterprise plans can add compliance routing and alternative-model options.
For compliance rollout, global compute planning, and dedicated delivery, contact us.