# AI UGC Video Generator > Agent skill on puras — published by puras. An AI UGC video generator: turn a product brief or URL (plus optional product photos) into an authentic, handheld talk-to-camera UGC ad with a cast creator, naturally paced dialogue, and a full soundtrack — ambient room sound, a subtle music bed, and product foley — featuring your real product, no filming. Works for a physical product, an app, or a service. - **Skill path:** `puras/product-ad-studio/ugc-video` - **Skillpack:** Product Ad Studio (`product-ad-studio`) - **Skillpack ID:** `39903015-da3d-49a2-b549-9256ef164430` - **Deployment version:** v58 - **Human page:** https://puras.co/skills/puras/product-ad-studio/ugc-video - **API base:** `https://api.puras.co` ## Description An AI UGC video generator: turn a product brief or URL (plus optional product photos) into an authentic, handheld talk-to-camera UGC ad with a cast creator, naturally paced dialogue, and a full soundtrack — ambient room sound, a subtle music bed, and product foley — featuring your real product, no filming. Works for a physical product, an app, or a service. ## Try it free AI UGC Video Generator has a free in-browser playground on its page (https://puras.co/skills/puras/product-ad-studio/ugc-video) — load an example or bring your own inputs and run it with a Google sign-in. No credit card, no subscription; runs are usage-based after the free try. ## Input schema ```json { "type": "object", "required": [ "brief" ], "properties": { "brief": { "type": "text", "maxLength": 2000, "minLength": 10, "description": "What it is, who it's for, what to land. URL works." }, "music": { "type": "string", "default": "auto", "maxLength": 120, "description": "Background music for the clip's native soundtrack. `auto` usually picks a quiet instrumental bed matching your brand's tone (lo-fi, soft acoustic, or light upbeat — always under the voice); raw-phone concepts like ASMR stay music-free. `none` always skips music for a raw phone-recording feel. Or type a style hint (e.g. \"lo-fi chill\", \"upbeat electronic\").\n" }, "captions": { "type": "boolean", "default": true, "description": "Burn word-synced karaoke captions onto the finished clip — the retention look for muted TikTok/Reels/Shorts feeds. Caption colors and styling are picked automatically to match the product's brand identity. Turn off for a clean, caption-free clip.\n" }, "last_frame": { "type": "image", "description": "Optional. A still the clip should end on (tail keyframe). When set, the render lands on this exact frame — e.g. an end card, or the last frame of a gameplay recording so the clip flows seamlessly into the next shot when stitched. Pairs with `first_frame` to interpolate from open to close.\n" }, "first_frame": { "type": "image", "description": "Optional. A still the clip should open on (image-to-video). When set, the render starts from this exact frame — e.g. the first frame of a gameplay recording so the clip flows seamlessly into the real footage when stitched.\n" }, "aspect_ratio": { "enum": [ "9:16", "1:1", "16:9" ], "type": "string", "default": "9:16", "description": "Output frame for the clip." }, "ugc_template": { "enum": [ "auto", "unboxing", "tryout", "interview", "tutorial", "product_review", "problem_solution", "grwm", "asmr_demo", "pov", "before_after", "listicle", "reaction", "storytime" ], "type": "string", "default": "auto", "description": "Which UGC format to shoot. `auto` picks the best fit." }, "creator_image": { "type": "image", "description": "Optional photo of the on-camera person, for a consistent creator." }, "product_images": { "type": "array", "items": { "type": "image" }, "maxItems": 5, "minItems": 0, "description": "Optional. Product photos that anchor the real product. Sharp, well-lit shots work best. If omitted, the product is rendered from the brief/research description." }, "creator_persona": { "type": "text", "maxLength": 300, "description": "Optional. Who's on camera and where. Empty = auto-cast." }, "duration_seconds": { "type": "integer", "default": 8, "maximum": 15, "minimum": 4, "description": "Target clip length. 8 s is the sweet spot; up to 15 s supported. Longer clips buy more action, not more talking — spoken lines stay short either way, so pick longer durations for demos and reactions, not bigger scripts.\n" } } } ``` ## Output schema ```json { "type": "object", "properties": { "video": { "type": "video", "description": "The rendered UGC clip (with native audio) — the playable video. The playground renders it with a