🎬 Global Launch: Dreamina Seedance 2.5 is now live!
Our most cinematic video model yet has arrived, and Dreamina AI is your official platform. Step into a new era of precision control and long-take storytelling:
· Native 30s videos in a single generation
· A new interactive editing experience: reshape any part of the frame
· Long video mode: up to 3 minutes, with consistency that holds throughout
· Dreamina AI plugins for Maya and Blender, built for film-grade production pipelines
· Up to 50 multimodal references, now with support for 3D white models and green-screen footage — bringing AI generation into real production workflows
· More true-to-life lighting and shadows, richer motion detail, and precise timestamp control down to 1 second
Dreamina Seedance 2.5 is more than a model. It's an upgrade to what you can create. Now open to Dreamina AI subscribers.
🕐 Available now across Southeast Asia, the Middle East, Africa, Europe, and South America. More countries and regions coming soon.
✨ Watch @yingjian355352 in action with the new model
⏳ Limited time: save up to 32% on credits per generation
🎁 Repost and comment – first 300 users get 1,000 credits in 12 hours
#DreaminaSeedance25 #DreaminaAI
187 credits saved on every continuation. Across a full film, that adds up fast.
Works on Dreamina with Seedance 2.0. Try it on other platforms and see if it holds.
For more AI video workflows, follow me: @MarinMethod
How to save credits in Dreamina when extending Seedance 2.0 videos:
Mine cost 510 credits per 15-second clip. Now I pay 323 for the same output.
One simple trick. Here's how: 🧵
The fix: cut your reference clip down to 4 seconds in your editor first.
Upload the 4-second version instead of the full clip.
Then generate your 15-second video like normal. Seedance only needs the last few seconds to pick up from.
New price: 323 credits.
From there I generated the clips in Seedance 2.0, tagged all the reference sheets, used the continue a video feature, added music and edited it all together.
If you want to see more AI video workflows, follow me @MarinMethod
I made a short AI film in Seedance 2.0 with 4 characters, 2 environments, 2 props and 1 vehicle.
Consistency held up across the clips.
Here are the prompt templates I used: 🧵
Goodbye Communities.
Thank you for everything.
The last 12 months we've given away $80,000 in cash prizes and $100,000+ in free access to AI tools.
Thank you to the Generative AI Community and the 250,000 people who contributed to making this place special.
Happy Horse-1.0 vs Seedance 2.0
Prompt:
A realistic shot of an old man in a cozy kitchen being jumpscared when his toaster launches the bread five feet into the air like a rocket. Handheld "home video" style capturing his genuine look of shock and the bread hitting the ceiling.A realistic shot of an old man in a cozy kitchen being jumpscared when his toaster launches the bread five feet into the air like a rocket. Handheld "home video" style capturing his genuine look of shock and the bread hitting the ceiling.
Prompt:
Cinematic medium-wide shot at night. No people in the scene. Two traditional five-story Japanese pagodas stand at equal height, facing each other across a gap. They are connected at their fourth story by a wide wooden bridge. The scene is lit only by a huge full moon. The camera is positioned close and elevated, at the same height as the bridge. The fourth and fifth floors of both pagodas fill most of the frame. The lower floors fall out of the bottom of the frame. The left pagoda has a wide wooden deck wrapping around its fourth floor. The deck is about three meters deep, clearly large enough for people to stand and walk on. It has weathered dark timber planking and a carved wooden railing at waist height. Behind the deck, glowing shoji paper doors line the inner wall of the building. Above the fourth-floor deck, the fifth floor of the left pagoda rises as a smaller tower with its own narrow balcony. The fifth floor is crowned with a traditional pyramidal tiled roof. The corners of the roof curve sharply upward into pointed eaves against the sky. A golden metal spire rises from the peak. The tiles are dark indigo with silver moonlight catching every ridge. The right pagoda is identical to the left. Same deck, same fifth-floor tower, same curved roof with upturned corners, same spire. Both pagodas are perfectly symmetrical in scale and design. A wide wooden bridge connects the two fourth-floor decks across the center of the frame. The bridge is about two meters wide, clearly walkable. It has heavy timber beams, carved wooden railings on both sides, and a gentle arch in the middle. The wood is weathered and dark. Moonlight catches the top edge of the railings in clean silver lines. The foreground has soft out-of-focus dark leaves and thin branches of a tree in the bottom corners of the frame. The leaves are nearly black silhouettes. They suggest the camera is hidden in a nearby tree at the same height as the bridge. The background shows a massive pale silver full moon. The moon is framed perfectly in the sky between the two pagoda towers, directly above the bridge. The night sky is deep indigo with a faint dusting of stars. Below the pagodas, rolling forested mountains fade into cool blue haze. Thin clouds drift between the peaks. Lighting comes entirely from the moon, behind the pagodas. Strong rim light traces the upturned roof corners, the bridge railings, the deck edges, and the wooden brackets. Cool silver highlights dominate the scene. Warm gold accents glow only on the spires and faintly through the shoji paper panels. Soft haze catches the moonlight in the air. The color palette is deep indigo and midnight blue, with silver highlights, warm gold accents on the spires, and near-black silhouettes in the foreground tree. The mood is high contrast, moody, and painterly. The scene has no lanterns, no banners, no modern elements. Style: shot on 35mm anamorphic lens, medium-wide composition, deep focus with soft foreground blur on the leaves. Kurosawa-inspired night grade with rich indigos, silver highlights, and crushed blacks. Subtle film grain. Photorealistic and cinematic. 16:9 aspect ratio. Make the aspect ratio 16:9
Fix for the noise/artifacts in GPT Image 2:
Start a new chat every time you want to generate a fresh image. (Ideally, keep the chat thread under 5 images).
The difference is wild.
Comparison using the same prompt:
⬅️ Left: Longer chat ➡️ Right: Fresh chat
PROMPT:
[CINEMATIC SETUP] Arri Alexa 35 / Zeiss Master Prime 32mm / f2.0 / handheld with occasional hydraulic stabilization during wide shots. Desaturated nordic color grade — cold steel blues, amber firelight, deep shadow contrast. Primary lighting: overcast midday sun breaking through storm clouds, secondary: distant burning longhouse casting warm orange rim light from left.
Atmosphere: frozen Nordic battlefield, ankle-deep ash and snow, ravens circling overhead, breath misting in cold air.
Audio: sword impacts, chain rattling, fur and leather in motion, guttural breath, ground tremors — no music. Face stable, no deformation. [@ image1] is THE GREY JARL, a veteran Viking warlord, long silver-blonde hair whipping in wind, black fur cloak, Celtic-engraved leather armor, Vegvisir runic shield, one-handed battle axe. [@ image2] is THE STONE BERSERKER, reference for massive hulking enemy — bald, stone-cracked skin, glowing orange runic chest plate, battered iron pauldrons, wielding a massive spiked flail-mace on chain, bare sandaled feet.
0–1s: EXTREME WIDE / LOW ANGLE. The Grey Jarl stands center frame, axe low, fur cloak catching wind. Silence except distant thunder. Slow push-in from ground level. Sound: wind, distant raven call.
1–3s: OVER-SHOULDER WIDE. The Stone Berserker enters frame from right, each footstep cracking frozen earth, twice the Jarl's mass. Camera holds static. Sound: heavy ground impacts, chain dragging, low guttural growl.
3–5s: CLOSE-UP INTERCUTTING. Tight on Jarl's weathered eyes narrowing. Cut to Stone Berserker's glowing orange runes pulsing brighter. Jarl raises runic shield with a slow grinding motion. Camera: static macro shots. Sound: runes humming, leather tightening.
5–7s: MEDIUM SHOT. The Berserker swings the spiked flail overhead in a wide arc, chain rattling violently. The Jarl sidesteps with veteran precision, the spiked ball cratering the frozen ground inches away, erupting a shockwave of ice shards and ash. Camera: crash zoom backward on impact. Sound: devastating whoosh, ground explosion, chain impact, ice shatter. 7–9s: TRACKING SHOT — HANDHELD CHAOS. The Jarl charges low, driving his shoulder into the Berserker's midsection. The Berserker barely staggers, grabbing the Jarl's fur cloak and hurling him sideways. The Jarl rolls through snow and rises into a defensive crouch, axe raised. Camera: whip-pan following body trajectory, rack focus to recovery. Sound: collision thud, snow crunch, chain swinging, strained breath.
9–11s: MEDIUM CLOSE-UP. The Jarl feints left, then drives his axe blade hard into the gap between the Berserker's iron pauldron and chest plate. Sparks. The Berserker roars — orange runes flare violently across his torso. He swings his massive arm, backhanding the Jarl's Vegvisir shield, splintering its outer edge. Camera: slow-motion 50% speed on axe impact, then real-time on shield shatter. Sound: metal scrape, deep roar, wood splinter crack.
11–13s: WIDE DUTCH ANGLE. Both warriors circle each other, both wounded — Jarl gripping fractured shield, Berserker leaking dark smoke from the axe wound, runes dimming slightly. Steam rises from both figures. Camera: slow dolly orbit. Sound: labored breathing, chain clinking, distant fire crackle.
13–15s: CLOSING SHOT — HERO FRAME LOW ANGLE. The Grey Jarl drives his battle axe down in a two-handed overhead strike directly into the Berserker's glowing chest runes. The runes shatter outward in an explosion of orange light and stone fragments. The Berserker drops to one knee. The Jarl stands over him, ash and embers swirling around his fur cloak, chest heaving. Camera: slow push-in holds on Jarl's face — grim, unbroken. Sound: rune-shatter burst, heavy collapse impact, wind, single distant raven, slow fade to silence.