LlmCall.Media.Dialect.Seedance.Video (llm_call v0.1.0)

Copy Markdown View Source

Seedance video generation (BytePlus ModelArk content generation tasks).

Submit POST /api/v3/contents/generations/tasks, then poll GET /api/v3/contents/generations/tasks/{id}. The prompt and any reference images are content parts; ratio, duration, and seed are top-level.

Native knobs this canonical layer has no field for (resolution, watermark, generate_audio, camera_fixed, service_tier, return_last_frame) are reached through provider_options.seedance.

durations_s declares the range this codec accepts; the core has no model catalog, so a model-specific limit surfaces as a provider error.

A finished task reports its token usage (completion_tokens, total_tokens), which lands on LlmCall.Media.Response.usage. The video URL is valid for 24 hours, so it has to be fetched soon after the task completes.

Summary

Functions

Runs the shared media validation pipeline with the codec's own checks.

Functions

validate_request(request, config)

@spec validate_request(LlmCall.Media.Request.t(), LlmCall.Config.t()) ::
  :ok | {:error, LlmCall.Error.t()}

Runs the shared media validation pipeline with the codec's own checks.