FREQUENTLY ASKED QUESTIONS
TokenAAS questions, answered clearly
Direct answers for developers evaluating the TokenAAS API and model catalog.
Updated 2026-09-11
TokenAAS provides one managed API layer for selected text, image, and video models, with model groups, capacity-aware routing, usage records, and billing controls.
Frequently asked questions
Is the TokenAAS API OpenAI compatible?
TokenAAS supports OpenAI-compatible request patterns for applicable endpoints. Use the documented TokenAAS base URL, your API key, and an available public model ID.
Which model families are listed?
The catalog includes selected DeepSeek, Qwen, GLM, Kimi, MiniMax, and Doubao model families. The catalog is the source of truth because availability changes.
How is text usage billed?
Text models are normally billed using input and output token rates. The model catalog displays the configured unit and price.
How are images billed?
Image models use per-request or configured image-tier pricing. Supported sizes depend on the selected upstream model.
How are videos billed?
Video models use per-second pricing, with resolution tiers where configured. Video generation is asynchronous: create a task, poll its status, then download the result with authentication.
What do Available, Limited capacity, Unavailable, Disabled, and Coming soon mean?
Available has routable capacity; Limited capacity has reduced capacity; Unavailable has no eligible capacity; Disabled is administratively disabled; Coming soon is listed for planning and cannot be routed or billed.
Does every API key see every model?
No. A key belongs to a model group, and that group controls which public model IDs the key can use.
Where should I check incidents?
Use the public Status page for current service health and incident information.