FAQ

How is total cost calculated?

Total cost = model invocation fee + platform service fee. Different plans affect service-fee policy and service level. See Pricing for details.

How does the 3% service fee apply?

By default, a 3% platform service fee is applied. Pro and Enterprise plans may receive waivers or custom terms by agreement.

What are the main differences between plans?

Differences mainly appear in RPM, SLA, support response, and compliance extension capability. Starter fits fast launch; Pro and Enterprise target higher concurrency and availability needs.

Which settlement currency is used?

USD is the core currency for both top-up and consumption billing.

Are failed requests billed?

Only requests that complete successfully with valid results are billed. Incomplete failed requests are not counted as billable invocations.

What model identifier formats are supported?

Three formats are supported: model name (for example deepseek-v3.2), provider/model (for example alibaba/deepseek-v3.2), and TokenHub plan model name.

What is the difference between deepseek-v3.2 and alibaba/deepseek-v3.2?

The former lets platform routing choose within allowed scope, while the latter pins a specific provider. Choose based on reliability, compliance, and procurement policy.

Do text, image, and video share the same invocation format?

Different modalities vary in parameters and task semantics. Check Text Models, Image Models, and Video Models documentation by scenario.

Where do I create an API Key?

Create it from API Keys in console after sign-in. Keys are shown only once at creation, so store them securely immediately.

Can routing exceed my configured scope?

No. Region, provider, and model constraints set in user preferences are hard boundaries. Subsequent requests route only within those configured limits.

What if no provider is available in the selected region?

Requests may fail when no provider is available in the chosen region. You can adjust regional policy or contact enterprise support for compliance extension options.

How does the platform protect availability during service issues?

The platform uses global availability observation for automatic failover. Per request, backend attempts up to three providers to maximize business continuity.

What should I do during large-scale network incidents?

Wide-area infrastructure failures may cause short-term unavailability. Retry later and monitor platform announcements.

What is the public conclusion of your routing strategy?

Current routing follows “User Compliance Requirements → Provider Availability → Cache Consistency” to balance compliance control with maximum availability. See Routing Strategy.

When should I upgrade to Enterprise?

Upgrade is recommended when you need higher concurrency, stricter SLA, dedicated support, compliance extensions, or global deployment capabilities.

What compliance extensions are available in Enterprise?

Enterprise supports broader compliance governance, including compliance routing extension, alternative-model consulting, and global delivery capability assessment.

How can I contact Enterprise and compliance teams?

Submit your request via Contact Us and we will arrange expert follow-up; you can also email tokenhublink@gmail.com.

Billing & Plans