Product Video Studio

What Problem Does It Solve?

When you have a product hero image and concise selling points, and your goal is to quickly create a publishable "product narration video" for e-commerce or social media (especially vertical short-video platforms), the traditional editing and voiceover process is time-consuming and requires professional skills. This skill automates this pipeline: given a static product image and a script, it outputs a vertical video with voice narration, subtle animation, and a duration of about 12 seconds, ready for publication.

Core Capabilities & The Technical Pipeline

This skill orchestrates AI capabilities across image processing, speech synthesis, and video generation. Its core workflow is as:

1. Image Preparation & Transformation:
   • First, it checks the input product image's format and specifications.
   • It uses beatra.images.transform to create a vertical 9:16 aspect ratio "hero frame" from the image. If the original is already a clean vertical image, it's used directly.
2. Voice Synthesis:
   • Based on the product's visible attributes and traceable selling points provided by the merchant, it crafts a narration script of approximately 12 seconds.
   • It selects a suitable voice style via beatra.voices.list and synthesizes the voiceover using beatra.speech.synthesize, obtaining the real duration and MIME type.
3. Video Generation & Delivery:
   • The prepared hero frame and voiceover audio are submitted to a video generation model (e.g., beatra.videos.animate).
   • The system strictly differentiates between "paid generation" and "non-paid preparation" stages, providing explicit cost estimates.
   • The final deliverable is the processed video file, accompanied by the returned dimensions, duration, and billing information.

Scope of Application & Key Constraints

• Intended Scenario: Exclusively for cases where "the product is the main visual subject + narration is required + the deliverable is a video".
• Diversion to Adjacent Skills:
  • Only needs product motion effects (no voice) → Use the "Image-to-Video" skill.
  • Real person/digital human presents the product → Use the "Digital Human Narration" skill.
  • Only needs enhancement of static product images → Use the "Product Image" skill.
  • Only needs voiceover audio generation → Use the "Voice Synthesis" skill.
• Information Dependency: The video script can be written based on the product's visible shape, color, surface, and use case. However, factual information like material, specs, certifications, pricing, warranty, and promotions must be provided by the merchant; if missing, the script will use expressions that don't rely on this data.
• Paid Operations & Confirmation: The skill design involves multiple paid API calls, and modifying key parameters (like aspect ratio) after the first payment incurs a high cost. The workflow requires strict user confirmation at critical junctures (preparation complete, before video generation) and ensures sufficient account balance.