higgsfield-video-explainer

What problem does it solve? Producing a consistent narrated explainer video requires coordinating style references, voiceover generation, per-scene video clips, and final assembly, which is error-prone when done manually across separate AI generation tools. ## Core Features & Use Cases - Block-based pipeline: Splits a video into fixed 10-second blocks, generating one Seed Audio narration take and one Gemini Omni clip per block, then assembles them server-side with explainer_video. - Style consistency: Locks one universal style key image (from live CMS presets or custom style-reference images) and reuses it across every clip to prevent style drift. - Flexible output options: Supports mascot or faceless modes, 16:9 or 9:16 aspect ratios, multiple narration languages, and optional burned subtitles. - Use Case: Turn a researched topic or a personal story into a 3-minute animated narrated video by picking a visual preset, choosing a narrator voice, and letting the pipeline generate and assemble all 18 blocks automatically. ## Quick Start Ask the agent to make a narrated explainer video about your topic, then pick a style preset and voice when prompted.