---
name: ad-ugc
description: "A UGC ad with the recurring AI character — ElevenLabs v4 voice with tags (Hope in the historical test), Scribe word times, a GPT Image 2.5 start frame of the character with the product in a real place, MiniMax H3 Max reference-to-video 9:16 480P (character sheet, product sheet, start frame, voice) with 4 beats on the word times and jump cuts; own or licensed room tone and product sounds, following the source author's preference rather than FlashBoards policy. 14 s. Part of the ad reference set."
---

These creative references describe source experiments with external tools. Use your own supported account and tools; this download installs no automation. Settings and timings are experimental observations, not current FlashBoards availability. Use only media and voices you have permission to use.

# ad-ugc · the creator and the product

> **Alternative external route:** Seedance 2.5 reference-to-video (`bytedance/seedance-2.5/reference-to-video`, references as @Image1… / @Audio1…, 4–30 s, native sound) was preferred in the source experiments. The measured side-by-side comparison was a product commercial, not every format; compare it with the H3 method below for your material.

> **Source author's sound preference (9.10.2026), not FlashBoards policy:** avoid ElevenLabs Sound Effects (`/v1/sound-generation`, `fal-ai/elevenlabs/sound-effects`) for effects, room tone, ambience and jingles; the author found them strange and louder than the voices and preferred ElevenLabs for voices (TTS v4, Speech-to-Speech, Audio Isolation). Historical demonstrations below and in related guides still include ElevenLabs Music beds and an earlier Sound Effects ambience test; those are separate experiments, not a universal voices-only rule. For your own mix, use permissioned film sound (line-free gaps, Demucs without vocals), the video model's native sound or supported video-to-audio tools; keep effects under the voice.

1. **Script ≤ 15 s**: hook (a confession, a whisper) → the product in action (open, sip) → the benefit in her words → CTA ("link in bio") with a wink.
2. **Voice**: ElevenLabs v4 with your chosen voice; example tagged script: "[whispers] Okay, I need to confess something... [normal] … [excited] … [playful] …" → 13.8 s. Scribe v2 → word times.
3. **Start frame**: GPT Image 2.5 Flare edit, low, 1152×2048; picture 1 the character sheet (face, hair, clothes), picture 2 the product sheet; "as if she films herself at arm's length, the camera itself never seen", the product on the counter with the label to camera.
4. **Video**: `minimax/h3-max/reference-to-video`, 9:16, 480P, duration 14, enhance disabled; images [character sheet, product sheet, start frame], audio [voice]. JSON prompt: output (vertical UGC, hand-held, two jump cuts), references (Image 1 only the face/hair/clothes, Image 2 "exactly this can … every time", Image 3 the first frame), shots from/to on the word times with the line in quotes; rules: one can, the camera never seen, no phones, no subtitles. Voice shift 0.00 s.
5. **Sound**: the generation's voice + short `aecho` + room tone at −13 dB + the product sound (can crack) placed where it happens on screen.

The voice, tagged script, keyframes and timings above describe the external experiment, not delivered or required inputs. Choose your own permissioned voice and product references.