Purpose
Kinovi positions itself as "one API for video, image, and audio models." According to its official website, it aggregates multiple AI models in one place, allowing users to call models through a single studio interface and a REST API without writing glue code or building ComfyUI graphs, so they can spend time creating rather than integrating.
Main capabilities
- Model aggregation: The official website lists "featured models," covering multiple output types such as video, image, and audio, and states "every model in one place" and "multiple model providers behind one endpoint."
- Video models: The official website shows that Seedance 2.5 (ByteDance) is already available, and lists video model routes such as Seedance 2.5 Video Edit, Seedance 2.5 Video Extend, Seedance 2.0, Seedance 2.0 Mini, Seedance 2.0 Fast, and HappyHorse 1.0 (Alibaba).
- Image generation and editing: The official website states that images can be generated or edited from prompts and reference inputs, with pay-as-you-go rates, through the AI Image Creator entry point.
- Unified calling workflow: The official website explains submitting "model + model-specific inputs" to createTask, then polling recordInfo with the returned task ID.
- Shared balance: The official website states pay-as-you-go pricing, with Studio and the REST API sharing one balance, and the ability to compare credit packages.
How to use
The official website provides two approaches: first, the Kinovi Studio interface, which the official website says can be used without coding; second, the public API, which the official website says calls createTask with the model's "model + model-specific inputs" and polls recordInfo. The documentation includes examples and lists the current model IDs and rates.
Use cases
Suitable for developers, creators, or teams that want to use one set of interfaces to call video, image, and audio models in a unified way, and that care about model IDs, unit prices, and documentation examples; also suitable for workflows that need a studio interface for creation while integrating the API to produce tasks when needed.
Limitations and missing information
The official website page does not provide a complete list of specifically available models, the exact prices of credit packages, concurrency or rate limits, the specific terms of commercial licensing,the specific value of the number of interface languages. The model prices listed on the official website include multiple resolutions and reference video variants, and show differences from standard prices with labels such as "Save 46%"; the specific current catalog on the official website shall prevail.
