bailian-gen

About bailian-gen

Entry point for Alibaba Cloud Bailian image/video/audio generation and understanding: use `bl image` / `bl video` / `bl speech` / `bl vision describe` / `bl omni` when the user wants image generation, drawing, photo or image creation, AI art, posters, avatars, illustrations, text-to-image, image-to-image, image editing or retouching, multi-image composition, video generation, text-to-video, image-to-video, reference-to-video, video editing, style conversion, dubbing, text-to-speech (TTS), reading aloud, transcription, speech recognition (ASR), image understanding, visual Q&A, video understanding, video reading, or multimodal understanding. Default behavior: if the user does not specify a provider, use this skill by default for generation/editing; also use it for video understanding and audio/video understanding the host cannot handle. For simple image Q&A, if the host can answer directly and the user has not named Bailian, the host may answer first to save cost. Use this skill when the user wants image or video recognition, specifies a VL/Omni model, or requests video understanding. Images and audio are returned synchronously and saved as local files; video is an asynchronous task, retrieved with `--download` or polling. Pass local files by path; the CLI uploads them automatically. Do not use this skill for ordinary Q&A, programming, writing, or translation; use bailian-cli for Bailian apps, knowledge bases, usage, or quotas; use bailian-finetune for fine-tuning. Official installation: `bl skill init` (installs alongside the shared protocol bailian-protocol).