Made with Seedance 2.5 — one native 15-second generation, no editing.

The idea was a playful gym vlog with an old camcorder feel: shaky framing, soft tape noise, imperfect focus, and someone questioning every life decision during core day.

The prompt covers six shots, from “Why am I shaking already?” during a plank to “Same time tomorrow? No promises.” at the end.

Curious what you think: does this kind of imperfect home-video style make AI footage feel more believable?

Prompt:

"Create a 15-second playful gym vlog featuring MILA, a 20-year-old American brunette with a high ponytail, wearing a matching cobalt-blue sports bra and athletic shorts, white ankle socks, and no sneakers.

CAMERA: Handheld tape-camcorder feel. MILA holds the camera herself for selfie shots and props it on the floor or a mat for exercise shots. Include hand shake, misaligned framing, delayed focus pulls, clumsy zooms, and occasional face cut-off framing. The camcorder never appears on screen.

LOOK: Soft, slightly blurry tape quality, faint tape noise, bloomed highlights under gym lighting, flickering auto-exposure, muted contrast, and realistic skin tones.

STYLE: Playful, self-deprecating gym-vlog tone. Genuine strain mixed with humor, complaining about core work while still pushing through it. Quick cuts between handheld and propped-camera shots. Keep the energy light and funny.

SETTING: A gym mat area in the evening, with a nearby mirror wall, a water bottle on the floor, soft overhead lighting, and equipment in the background.

STORYBOARD:

1. (~2.5s, propped camera, medium shot) She lies on the mat, props herself up on her elbows, and sighs dramatically. MILA: “Core day. Pray for me.”
2. (~2.5s, propped camera, plank position) She holds a plank, arms shaking slightly, talking through gritted teeth. MILA: “Why am I shaking already?”
3. (~2.5s, medium propped shot, crunches) She performs crunches, complaining between reps. MILA: “This looked way easier online.”
4. (~2s, close-up insert, shallow depth of field) Her hands grip the mat edges during a leg raise. No dialogue, ambient gym sound only.
5. (~2.5s, handheld feel, leg raises) She finishes a slow leg raise set, flops back onto the mat, and laughs. MILA: “Yep—my abs just quit.”
6. (~3s, arm’s-length selfie finish) Still lying on the mat, she lifts the camera above her face and grins tiredly. The camera remains out of view. MILA: “Same time tomorrow? No promises.”

Share your thoughts about Seedance 2.5's realistic videos below!

I wanted to test how well Seedance 2.5 handles realistic, everyday content instead of cinematic AI videos.

So I created this 30-second Get Ready With Me (GRWM) morning vlog.

The goal was to make it feel like a real TikTok recorded on a smartphone — one continuous handheld selfie shot, natural lighting, imperfect camera movements, casual dialogue, and small everyday interactions.

What I focused on:

• Consistent facial identity throughout the entire 30 seconds
• Realistic handheld smartphone movement
• Natural dialogue and lip sync
• Hair adjustments, makeup, outfit selection, and walking
• Authentic room ambience without background music
• No cinematic effects or artificial beauty filters

I think these casual, everyday scenarios are an interesting way to test AI video models because even tiny inconsistencies can make the footage feel fake.

Prompt:

"Create a 30-second casual “get ready with me” morning vlog featuring the same young woman in her early twenties. She has natural everyday beauty, and her face and hair remain identical throughout the entire video. Camera & visual style One continuous handheld front-facing smartphone selfie vlog, filmed entirely by her with the phone held in her own hand at arm’s length. No third-person camera, tripod, cinematic shots, professional camera movement, or external filming. Make it look like an authentic modern smartphone recording: - Mild handheld movement and natural arm motion - Small framing imperfections - Occasional autofocus adjustments - Natural exposure fluctuations - Realistic front-camera lens distortion - Soft morning window light - Natural smartphone colors - No beauty filter, skin smoothing, or cinematic color grading - Same woman, same face and hair throughout - She holds the phone herself at all times - No subtitles, logos, watermarks, captions, or on-screen text Scene She starts in her bedroom shortly after waking up, holding the phone toward herself with a sleepy smile. She looks into the camera and says: “Okay, I have approximately ten minutes to look like I have my life together.” She laughs, briefly turns the camera toward her closet, then returns it to her face. She casually fixes her hair, applies a small amount of makeup while occasionally looking into the phone camera, chooses an outfit, and walks around her room. Include tiny natural moments: adjusting her hair, checking herself in a mirror, realizing she forgot something, laughing, and briefly looking away from the camera. She grabs her bag and heads toward the door. Before leaving, she looks into the camera and says: “Okay, that’s good enough. Let’s go.” She gives a quick smile and lowers the phone slightly as the clip ends. Audio Diegetic sound only. Include quiet bedroom ambience, fabric movement, footsteps, drawers opening, subtle street ambience, and natural room tone. No background music. The final result should feel like an authentic, imperfect morning vlog recorded casually on a modern smartphone."

What do you think? Could this pass as a real GRWM vlog, or can you immediately tell it's AI-generated?

Model: Seedance 2.5
Duration: 30 seconds

Curious to hear your thoughts!