PixVerse C1 Video Generation

Problem It Addresses

pixverse-c1 targets workflows that call the PixVerse C1 video model to generate video output: provide a prompt, or upload an image, first/last frames, or reference assets, then retrieve downloadable video results. It does not change the underlying model capability; it wraps the dLazy hosted API as a CLI-style invocation path so scripts or agents can run text-to-video, image-to-video, first/last-frame-to-video, and reference-to-video generation with structured parameters.

How It Works

The skill behaves as a thin client over the dLazy CLI. Key steps include:

• Authenticate with a dLazy API key, stored in the user config directory or supplied per invocation via DLAZY_API_KEY.
• Send parameters such as prompt, image, video, or audio to api.dlazy.com for inference.
• When local file paths are provided, the files are uploaded to dLazy media storage so the model can read them.
• The API returns hosted output URLs; for asynchronous jobs, a generateId can be used to query task status later.

Boundaries And Caveats

It depends on an external cloud service and local CLI, not pure on-device inference. Pay attention to API key scope, account balance, file upload destination, and service terms. Common failure modes include unauthorized access, missing required parameters, local file read errors, insufficient balance, and backend asynchronous task failures. If the result returns insufficient_balance or unauthorized, guide the user to top up credits or configure an API key instead of blindly retrying.