Dreamina AI was developed by CapCut, a subsidiary of ByteDance (company behind TikTok). It offeres a host of features form text to image, to image to video, but most interesting one is it's lip sync.
Dreamina's OmniHuman-1 is built to take one still image plus a voice track and spit out a short clip that actually feels human. We’re talking spot-on lip sync expressive gestures and even full-body movement if your source image supports it.
Just drop in an image and audio (or text and select one of the pre-built voices) and you’ll get a 15-second clip that feels personal polished and ready to post. Want vertical? Horizontal? 4:3 nostalgia-core? It’s got you.
And it’s not just for face crops. The model works with full-body shots too and keeps proportions tight so gestures don’t go off the rails.
It also does some animal lip sync, though success rate may vary (check examples below).
Creators can roll out quick clips that look like studio-quality puppet animation. Marketers can humanize brand messages without booking talent. Teachers could make explainer avatars. Game devs and VR creators get a boost too since it gives you a fast way to prototype lifelike characters without a modeling budget.
Right now you’re capped at 15 seconds and the AI handles all the direction. You can’t feed it a prompt like “look left” or in any way drive the animation, you get what it gives. But it's pretty good.
Tags
Freemium Proprietary License Web-based#Creative AI Suites
Educators and TrainersCreative ProfessionalsContent CreatorsMedia and Film MakersMarketing and Branding SpecialistsVoice and Audio ProfessionalsDevelopers and Tech CreatorsNonprofit and Advocacy CreatorsSmall Business OwnersEntertainment and Performance ArtistsProfessional Content Creators
This list may not be exhaustive as new models keep dropping and are added to platforms all the time.
Prompt:
A timelapse begins from a serene dark-turquoise ocean reflecting moonlight symmetrically framed by mountains on both sides. The camera locks on the moon as an eclipse begins. The stars swirl subtly across the bright starry sky with darkish clouds. VFX Text Appearance: As the moon darkens, its glowing edges form a halo that flickers and dances. Shadows creep across the landscape. Reflections dance in the ocean water. The word “ECLIPSE” emerges in front of the eclipsed moon stretching slightly outside its radius. Cinematic, futuristic title text spelling “ECLIPSE”, centered in the night sky. The font is a thin, glowing, neon-like sans-serif with extended horizontal lines. Smooth, rounded edges. Letters slightly spaced out, minimalist, sci-fi look. The glow subtly matches the moonlight. Highly stylized, modern typography with a soft reflection on a dark surface. Just before totality, the entire scene plunges into darkness. A faint corona shoots around the dark moon.
Static shot of a woman standing on a mountain. She turns towards the camera smiling, and begins to turn pose confidently and slightly playfully like a model. In a series of seamless transitions, her outfit shifts: a sparkling mini dress => into => denim shorts overalls; grey blazer => into ... => a burgundy-red bomber; denim overalls => into => straight torn light-blue jeans. Wardrobe swaps are cut-style jumps, immediate and total. all while the backdrop of majestic mountains and blooming flowers enhances the scene's beauty. Hot air balloon behind her is branded "www.AIcreators.tools" is slowly floating right. The lighting is warm and inviting, creating a dreamy atmosphere, with a cinematic style .
Good model for this type of prompt. Clothes shift instantenously and now there's also a cinematic 'whoosh' sound accompanying the magical transition. Camera fails to stay static though.
A confident 25-year-old woman stands before a vibrant yellow wall painted with the phrase “AI creators tools.” She wears a chic purple jacket and embellished jeans, a black backpack in her hand. In a calm, rhythmic motion, she begins modeling her outfit with poise — stepping closer in fashion-forward pose while looking confidently at the camera, projecting strength, then wearing backpack on her shoulder and reatreating and adjusting her jacket, turning around herself. The camera remains almost still, holding focus on her, while a gentle parallax effect and soft depth-of-field blur shift subtly as she moves. The diffused lighting glows warmly, reflecting delicate highlights off her clothes, giving the entire frame a rich, fashion-editorial quality in crisp 4K.
Confident modeling, appropriate soundtrack and ok handling of the backpack (its handles change a bit but at least it doesnt hang in the air by itself).
High-quality, cinematic footage of a dialogue with humorous vibe. Two dogs: a Miniature Schnauzer on the left and a chocolate-brown Labrador on the right sit in a professional podcast studio.
On the walls, there are framed pictures of various dogs with vane and noble expressions and a large certificate-award reading "Certified Good Boy*" in bold and below, smaller "*Self-Certified".
The Miniature Schnauzer dog on left starts by saying enthusiastically while turning its head slightly towards the Labrador: “Mine talks to the TV like it can hear him?” - then turns head back straight.
The Labrador on right briefly glances at Schnauzer & replies cheerfully: “Mine argues with it. ARGUES. With a rectangle.”
In the end both dogs chuckle & laugh together.
Sweet. As long as you give the model enough seconds for your dialogue it renders it correctly with multiple speakers (even dogs in this example). Do not use all-caps for emphasis, it compells the model to spell the word.
A smoky backroom where four capybaras dressed in 1940s gangster attire sit around a poker table under a brass hanging lamp. Cigars smolder in the mouths of the two of them: capybara on the right and capybara with slicked fur in the back, tucked snugly beside their large front incisors, poker chips clatter, and a portrait of a glamorous capybara in a silky dress hangs slightly askew on the dark wooden wall. The camera cuts to a close-up of one capybara with slicked fur and a thick cigar - his eyes narrowing with suspicion as he studies the cards and his opponents through the haze saying "I smell a rat". The camera lingers on the glowing cigar tip, then cuts back to a wide shot of the table as the capybaras exchange cards, chips sliding across the felt under the golden light, the tension thick and cinematic.
Great result. Background music, camera cuts, consistent players, capybara's voice... I've given it a wrong start image version, with the cards being revealed, it's not the Seedance's fault
A couple sits at a small white iron table outside café. They hold hands and look at each other. The shot stays steady with a light film-like grain. It starts focused on the couple, then shifts to the back. the woman says, “He’s in New York till Friday, darling.” The man in blue shirt who is sitting at the table says, “So I can have you all to myself.” Meanshile silent man with a suitcase slowly walks into view, rack focus shifting to his shocked face. That changes the mood fast. The street has striped awnings, café chairs, and a busy but quiet flow of people. You hear street sounds, some footsteps, and soft clinks of dishes. No traffic noise or music.
A confident 25-year-old woman stands before a vibrant yellow wall painted with the phrase “AI creators tools.” She wears a chic purple jacket and embellished jeans, a black backpack in her hand. In a calm, rhythmic motion, she begins modeling her outfit with poise — stepping closer in fashion-forward pose while looking confidently at the camera, projecting strength, then wearing backpack on her shoulder and reatreating and adjusting her jacket, turning around herself. The camera remains almost still, holding focus on her, while a gentle parallax effect and soft depth-of-field blur shift subtly as she moves. The diffused lighting glows warmly, reflecting delicate highlights off her clothes, giving the entire frame a rich, fashion-editorial quality in crisp 4K.
A delicate butterfly flutters into frame and lands on a purple flower, placed slightly off-center with a white middle, captured in extreme macro detail against a soft, greyish blurred backdrop. The fragile wings shimmer under faint sunlight as the camera holds steady. Then a rack focus sharpens the background — heavy tank barrels emerge, dust rising, soldiers’ legs rushing past as shouts echo through the street. war-torn street engulfed in rubble, smoke, and fire. a stark, cinematic portrait of serenity and destruction.
Dreamina's Avatar Turbo had problems with this one. It's odd it even accepted this image to work with, usually if it thinks it can't spot the mouth it refuses to start the job.
A symmetrical, two-story blue house with white trim and a bright red door stands perfectly centered in the image The background features a bright blue sky with scattered cumulus clouds, adding depth and contrast. The house is the dominant central subject, with the tulip-lined path creating strong leading lines toward it. The camera dolly in slowly as if this is an ordinary suburban home. Suddenly, a massive human hand enters the frame, pulling open the house which turns out to be a dollhouse. On either side, tall hinged panels swing open—each painted pale blue to match the house’s exterior. These panels, with cut-out windows perfectly aligned with those on the side walls, reveal themselves as the front façade, now split like cabinet doors to expose the interior.
Push into the open dollhouse-style cross-section. Inside, tiny but realistic man in the top-left bedroom jolts upright in bed, startled, yells out from fright then stops.
Lite is 'almost there'. It can't quite figure out how the dollhouse opens but tries to make the next best thing by just touching the house and pusing into an open door.
👀 Dreamina's now got Multi-Frame mode for Seedance Lite and they're 7 frames.
feature
May 31, 2025
Seedance 1.0 lite is a lightweight video generation model that supports text-to-video and image-to-video creation, offering 5 or 10-second clips in 480P and 720P resolution. Despite having fewer parameters, it is claimed to deliver high-quality output with faster generation speeds. Has improved instruction-following and enhanced control over details like facial expressions, clothing, and intensity of actions.
model
April 9, 2025
Can be used for free right now as part of a promotion, for an unspecified length of time.
