Videos mit KI generieren – und das mit Open Source? WAN 2.2 von Alibaba hat da etwas vorgelegt, was überrascht. Das Modell beeindruckt nicht nur mit der Videoqualität, sondern auch bei der Bildgenerierung. Wir haben uns in diesem Beitrag die Videogenerierung mal genauer angeschaut und zeigen dir ein paar Beispiele, wie das Ganze aussehen kann – inklusive der passenden Prompts.
Inhaltsverzeichnis
Die wichtigsten Fakten zu WAN 2.2
Entwickler: Alibaba DAMO Academy
Lizenz: Apache 2.0 (kommerzielle Nutzung erlaubt)
Release: Ende Juli 2025
Modelle:
- T2V-A14B (Text zu Video)
- I2V-A14B (Bild zu Video)
- TI2V-5B (Hybrid)
Wie sich WAN 2.2 von anderen Video-KI-Modellen abhebt
Der wichtigste Unterschied bei WAN 2.2: Es ist Open Source. Das bedeutet, dass der Quellcode und die Modelle frei zugänglich sind. Du kannst sie einsehen, anpassen und kommerziell nutzen. Keine Blackbox, keine restriktiven Lizenzen.
Modelle wie Kling, Veo 3 oder Sora funktionieren nur innerhalb der Plattform des Anbieters. Du bist an deren Regeln gebunden. WAN 2.2 dagegen setzt auf Offenheit.
Das bringt dir volle Kontrolle über das Modell und mehr Flexibilität bei eigenen Anpassungen. Auch mit Consumer-Hardware lassen sich gute Ergebnisse erzielen. Durch die Community entstehen laufend neue Verbesserungen und Erweiterungen.
So wird WAN 2.2 zu einem offenen System, das sich klar von den geschlossenen Plattformen der großen Anbieter abgrenzt.
Was bei WAN 2.2 auffällt
Gute Prompttreue
Komplexere Beschreibungen werden besser verstanden als bei vergleichbaren Modellen. „Warmes Gegenlicht“ wird nicht automatisch zu einem Sonnenuntergang.
Prompt
Photorealistic, elegant 8-second scene at a European sidewalk café with a subtle deadpan-comedy twist. NO AUDIO: render completely silent (no ambience, music, dialogue, or SFX).
CHARACTER — A stylish woman in her early 30s with light skin and a neat low chignon of light brown hair. Black cat-eye sunglasses, beige trench coat over camel trousers, black block-heel pumps, minimal gold stud earrings, natural makeup. Calm, composed posture.
SETTING — Late morning golden sunlight (5600K) with dappled leaf shadows. Small round white-marble bistro table with an espresso cup and saucer; classic rattan café chairs in green-and-cream weave; potted shrubs; soft reflections in the café window behind her. Gentle breeze that lifts a few loose hair strands. Background pedestrians soft and out of focus.
ACTION —
0–2s: Medium shot at eye level, simulated ARRI Alexa Mini look, 50mm lens at f/2.8; a gentle 2-second dolly push-in from camera right as she reads a broadsheet newspaper.
2–3.5s: She turns one page smoothly; the breeze subtly flutters a few strands of hair.
3.5–6s: A waiter passes behind her left-to-right. He wears a crisp white shirt, black vest and apron — plus a bright pink tulle ballet tutu around his waist; black dress shoes. He walks professionally, expression neutral, the tutu swishing clearly; he carries a tray with two white cups and a bottle of sparkling water. Keep him slightly out of focus but unmistakable.
6–8s: The camera settles to a tighter medium (about 65mm equivalent). She briefly glances in his direction with the faintest amused smirk, then returns to the newspaper; hair moves lightly in the breeze again.
MOOD & STYLE — Chic editorial realism with understated humor. Warm, slightly desaturated palette (beige, olive green, ivory). Soft sunlight with a 3:1 key-to-fill ratio and gentle rim light. No exaggerated reactions; keep gestures subtle.
CAMERA & LENS — Start medium at eye level; 2-second dolly push-in; maintain shallow depth of field for elegant background blur. Simulate 50mm → 65mm push (no zoom artifacts), subtle parallax as the waiter crosses; end on a steady hold. Aspect ratio 16:9.
Prompt
Photorealistic, stylish 8-second portrait scene with a cool, modern editorial vibe.
CHARACTER — Young man (early 20s), fair-to-light skin with freckles, medium-slim build, dark brown curly hair of medium length, clean-shaven. Calm, confident demeanor. Wardrobe: black fitted turtleneck. Distinctive accessory to use: sporty wraparound sunglasses with a matte white frame and mirrored lenses.
SETTING — Intimate indoor corner with textured red-brown brick walls forming a shallow V behind him. Cool daylight-balanced look (~5200K) with soft key from camera-left (large diffused source), gentle negative fill on camera-right, and a subtle hair rim for separation. Background stays quiet and uncluttered.
ACTION (timed for 8s) —
0–2s: Medium close-up at eye level, 85mm portrait lens, f/2.0. He faces 3/4 left, gaze off-frame; slow 2s dolly push-in, catching tiny hair movement and breathing.
2–3.5s: He turns his head toward the lens and establishes eye contact; micro smile.
3.5–4.5s: He gives a single playful wink (right eye), holding eye contact.
4.5–6.5s: He raises sporty sunglasses from just below frame and slides them on smoothly; small head tilt to settle the fit; mirrored lenses catch a soft highlight.
6.5–8s: He steps out of frame to camera-right with confident ease. Camera makes a gentle 10–15cm pull-back and holds on the empty brick corner for a beat.
MOOD & STYLE — Understated, cool, slightly moody elegance. Neutral-to-cool color grade with moderate contrast; clean skin tones; minimal grain. Keep gestures subtle and natural; no exaggerated expressions.
CAMERA & LENS — ARRI Alexa Mini / 4K look. Start at 85mm (portrait compression), shallow depth (f/2.0), eye-level angle. Smooth micro-dolly push-in, then micro pull-back at the end; no whip pans, no Dutch tilt. Eye autofocus on the nearer eye; brief rack focus to background as he exits.
NOTES — Keep framing vertical-friendly with headroom for a 9:16 crop; avoid brand logos and on-screen text. Aspect ratio 9:16 portrait unless specified otherwise.
Prompt
Photorealistic 8-second cinematic shot of a rocket launch, capturing the power and scale of liftoff.
CHARACTER — Large modern orbital launch vehicle with white body, black nosecone, side boosters, and visible mission decals. Clean, realistic detailing with no fictional logos.
SETTING — Open launch pad at dawn, faint pre-sunrise glow with scattered clouds. Tower structure beside rocket, grassy field in foreground, distant treeline in soft haze.
ACTION —
0–2s: Locked-off medium-wide at 50mm, tripod-stable. Rocket engines ignite with brilliant orange-white flame; thick white smoke and vapor pour outward across the pad.
2–5s: Slow vertical camera tilt up to follow the rocket rising; bright exhaust plume extends down, turbulence rippling. Debris and steam swirl at base.
5–7s: Continue tracking as rocket clears the launch tower; background clouds lit by engine glow; boosters vibrate with thrust.
7–8s: Hold on rocket climbing toward the clouds, plume narrowing. Smoke continues to billow across the ground below.
MOOD & STYLE — Epic, powerful, inspiring. High contrast between bright engine flare and soft dawn tones; slight HDR effect for detail in highlights and shadows.
CAMERA & LENS — Simulated RED Digital Cinema 8K; 50mm for initial pad shot, smooth tilt and follow; deep focus (f/11) for sharp rocket and plume. Shutter ~1/125 for realistic motion blur.
LIGHTING — Natural dawn ambient with intense artificial engine glow dominating. Subtle rim from rising sunlight breaking the horizon.
NOTES — No fictional elements or unrealistic behavior; realistic launch physics and exhaust movement. Aspect ratio 16:9 unless otherwise specified.
Natürliche Bewegungen
Besonders bei Personen und Flüssigkeiten wirkt die Animation überzeugend. Haut, Haare und Wasser bewegen sich realistisch.
Prompt
Photorealistic, intimate 8-second bathtub portrait with the water as the star: a woman submerges, resurfaces, and wipes water from her face as ripples and petals move naturally.
CHARACTER — Woman mid/late 20s, light skin, relaxed expression, wet brunette hair slicked back; no visible nudity (frame shoulders-up). Minimal makeup.
SETTING — White porcelain tub filled with opaque milky bathwater; scattered deep-red rose petals. Soft window light from camera-left (~4200–4600K) with gentle rim highlights; faint steam. Clean, uncluttered bathroom.
ACTION (8s) —
0–2s: Top-down overhead shot (90°), tight medium at 35mm, f/3.2; tiny surface ripples and drifting petals in focus. Micro 10 cm slider push-in.
2–3.5s: She takes a breath and slowly sinks beneath the surface; small bubbles rise; petals part.
3.5–6s: Speed ramp to 60% slow motion as she resurfaces; water sheets off forehead and nose, droplets trailing from eyelashes and hairline.
6–8s: She brings both hands up, wipes across face and over crown; water streams between fingers, droplets fall back, concentric ripples radiate and swirl the petals.
MOOD & STYLE — Calm, spa-like, sensory; emphasize texture of water, translucence, and gentle movement. Soft contrast, creamy whites, rich crimson petals; avoid eroticization.
CAMERA & LENS — ARRI Alexa 35 look; overhead rig; 35mm → slight push; shutter 180°. Eye focus on surface plane with smooth micro rack to eyes as she emerges. Medium shallow DOF for crisp droplets with soft edges.
LIGHTING — 3:1 key-to-fill; diffused key from left, subtle negative fill right; add small specular kicker to make droplet highlights sparkle.
NOTES — No logos/text. Aspect ratio 9:16 portrait unless specified; safe for 16:9 crop.
Prompt
Photorealistic 8-second kitchen scene with a magical twist: a dropped egg bursts into pink glitter dust, transforming the entire kitchen into a sparkling pink room.
CHARACTERS — Woman mid-20s, East Asian, long dark hair in loose ponytail, oversized light-blue shirt, jeans, playful. Man late-20s, East Asian, short dark hair, charcoal shirt, cream-and-black striped apron, holding wooden spoon and nonstick pan.
SETTING — Modern white kitchen, evening warm-neutral light. White counter island, bowls, oil bottle, gas burner with black pan. Mid-scene transforms into “pink glitter room” with glossy blush-pink counters, glittery magenta backsplash, shimmering floating particles.
ACTION —
0–2s: Medium two-shot, he stirs while they smile at each other.
2–3s: She drops an egg; instead of yolk, radiant pink glitter bursts out in slow motion.
3–5s: Rack focus to their amazed, laughing faces as glitter swirls and the room’s surfaces turn pink and glittery.
5–7s: He flicks the spoon, sending sparkles; she claps in delight.
7–8s: Pull back to reveal the fully transformed glitter room as sparkles float in the air.
MOOD & STYLE — Romantic-comedy whimsy meets magical realism. Start with natural colors, transition to saturated pink-magenta palette with pearlescent highlights and gentle bloom.
CAMERA — ARRI Alexa Mini look. 35mm lens, whip-tilt to counter, slow motion on glitter burst, rack focus to faces, short dolly push-in, ending with pull-back reveal. Medium depth of field for sharp faces and sparkling particles. Aspect ratio 16:9 (safe for 9:16).
Hoher Detailgrad
Auch bei Nahaufnahmen bleiben Texturen und Oberflächen scharf und detailliert.
Prompt
Photorealistic macro nature scene, 8 seconds: a honeybee approaches lavender, lands, and feeds — capturing delicate pollen detail.
CHARACTER — Single western honeybee (Apis mellifera), fuzzy amber thorax, striped abdomen, translucent wings with subtle iridescence, dark eyes. Pollen baskets initially sparse, then dusted with pale yellow grains after contact.
SETTING — Sunlit garden with pale wooden fence bokeh. A slender lavender spike (soft purple florets with tiny orange stamens) stands slightly swaying in a light breeze. Cool-clean late morning light ~5600K; airy, high-key background with creamy whites and pastel greens.
ACTION (timed for 8s with gentle speed ramps) —
0–2s: Begin on the lavender spike in sharp focus, extreme close-up at 100mm macro, f/4. Slow 2-second slider push-in. Background creamy and bright.
2–4s: From frame left, the bee enters in 50% slow motion, hovering; micro parallax as she aligns. The stem quivers as she touches down near the upper florets; wings blur then fold.
4–6s: Tighten to an ECU (macro rail push a few centimeters). She unfurls her proboscis into a floret and drinks; hind legs brush anthers, knocking loose sparkling pollen motes that float in backlight. Focus tracks to the proboscis and pollen grains.
6–8s: She sidesteps to the adjacent floret, grooming a hind leg over the pollen basket; wings flick once. Hold on the shimmering pollen and subtle lavender sway as she continues feeding.
MOOD & STYLE — Gentle, mesmerising, nature-documentary intimacy. Clean, slightly cool whites, soft contrast, pastel lavender tones; emphasize translucency of wings and the floating pollen.
CAMERA & LENS — ARRI Alexa Mini (macro profile) or equivalent; 100mm macro lens; 120 fps capture for slow-motion moments; 1/180 shutter. Eye-level macro angle at ~45° to the spike; ultra-stable micro slider/rail. Shallow DOF with smooth focus pulls; no zoom artifacts.
LIGHTING — Natural soft sun with light negative fill from camera-right; subtle rim/backlight to make airborne pollen sparkle.
NOTES — Keep framing clean with no other insects; avoid text or logos. Aspect ratio 16:9 unless specified; allow safe crop to 9:16.
Drei Anwendungsbeispiele für Marketing
1. Getränke-Werbung mit Slow-Motion
Ein Glas steht auf Marmor, eine goldene Flüssigkeit fließt langsam hinein. Sonnenlicht erzeugt schöne Reflexionen. Klassisches Werbe-Setup, das mit WAN 2.2 gut funktioniert.
Warum es klappt: WAN 2.2 kann Flüssigkeiten und Lichtbrechung sehr gut darstellen.
Prompt
Ultra-photorealistic commercial close-up of freshly squeezed orange juice being poured into a clear, faceted crystal tumbler packed with large, transparent square ice cubes. Sunlit marble countertop near a window at late morning; warm backlight at 5200K with a 3:1 key-to-fill ratio, gentle bounce fill from the front, subtle rim light to catch condensation. Fine droplets sparkle on the glass; a halved Valencia orange and a wooden citrus reamer sit softly out of focus in the background. No humans.
ACTION (8 seconds, slow-motion aesthetic):
• 0–2s: A smooth stream of vivid orange juice enters from top right, striking the ice and forming a crisp splash crown; tiny micro-pulp particles swirl.
• 2–5s: The camera performs a slow 120-degree arc at macro distance, revealing refractive highlights through the glass facets; tiny bubbles drift upward around the ice.
• 5–8s: A gentle dolly push-in to an extreme close-up of condensation beads sliding down the chilled glass as the liquid level settles; finish on a bright, thirst-inducing glint.
MOOD & STYLE:
Refreshing, energetic, premium beverage ad; saturated oranges against cool whites; high-contrast specular highlights with controlled reflections (use a polarizing filter). Clean, modern color grade; crisp commercial polish; “ice-cold” vibe emphasized by visible frostiness and condensation.
CAMERA & LENSES:
ARRI Alexa Mini look; 100mm macro for the arc; 50mm for the final push-in. Shallow depth of field at f/2.8, 180° shutter feel; tripod + micro-slider for smooth motion; level horizon (no Dutch tilt).
2. Sport/Lifestyle in der Stadt
Eine Läuferin bewegt sich durch morgendliche Straßen. Die Sportkleidung bewegt sich natürlich mit, warmes Licht fällt auf die Gebäude.
Warum es klappt: Natürliche Bewegungsabläufe sind eine Stärke von WAN 2.2.
Prompt
Photorealistic urban running commercial, 8 seconds.
CHARACTER & WARDROBE (focus on apparel):
Athletic woman in her late 20s, lean runner’s build; face mostly out of frame to emphasize clothing. Outfit: matte black high-waisted compression leggings with reflective piping and breathable mesh calf panels; seamless moisture-wicking sports bra; ultralight ripstop windbreaker in vibrant coral with micro-perforations and soft elastic hem; knit running shoes (engineered mesh upper, white cushioned midsole, subtle orange accents); black sports watch; no visible logos. Realistic fabric physics: natural stretch, subtle creasing at knees, ripstop jacket flutter in the slipstream, zipper pull bouncing, no clipping.
SCENE SETTING:
Modern city at sunrise (golden hour, ~4300K), long shadows, warm light kissing glass-and-concrete facades; tree-lined boulevard with light morning haze; puddle streaks and soft reflections on asphalt; sparse commuters in the distance, out of focus. Clean, premium lifestyle vibe.
CAMERA & LENSES:
Gimbal-based tracking, dynamic but smooth. Start on a low 15° angle with a 24mm wide for speed and environment; transition to a 50mm medium close-up for apparel detail. Shallow depth of field (f/2.8), natural motion blur (180° shutter feel). Subtle polarizer to tame window reflections; neutral density to hold highlights. ARRI Alexa Mini LF look.
ACTION (time-coded for 8s):
• 0–2s: Low-angle tracking beside her stride; crisp heel-to-toe footfall, leggings’ reflective piping catching warm sun; jacket hem flutters.
• 2–6s: Side-profile medium close-up on torso: coral windbreaker ripples, fabric stretches over shoulders; zipper pull and drawcords bounce; micro sweat sheen forms at collarbone; building parallax sells speed.
• 6–8s: Dolly push-in to a hero detail of the leggings’ texture and reflective seam as she accelerates past camera into a gentle sun flare; end on a clean freeze of the apparel, sharp and aspirational.
MOOD & STYLE:
Energetic, premium sport/lifestyle; golden highlights with cool city shadows (subtle teal in mids), high micro-contrast, clean commercial grade. No brand marks. Emphasize crisp textiles, breathable performance, and movement.
NOTES:
Keep the runner centered from waist to knees for most of the shot to prioritize apparel; maintain realistic cloth sim with slight wind from camera-left; no voiceover, no captions, no logos, no text on screen.
3. Produktpräsentation mit Rotation
Ein grünes Pesto dreht sich in einem minimalistischen Setting. Alle Details und Reflexionen werden sichtbar.
Prompt
• The jar rotates a full 360° on an invisible turntable over the entire 8 seconds at constant speed; perfectly stable, no wobble.
• The camera remains mostly locked with a subtle 10% dolly push-in to amplify detail and reflections.
• At ~4 seconds, a gentle 5° tilt-down reveals the top surface swirl and oil sheen while keeping center-framed composition.
CAMERA & LENSES:
ARRI Alexa Mini LF look; 85mm macro prime; aperture f/8 for crisp detail throughout; 180° shutter look; low ISO for clean blacks. Tripod + micro slider for the push-in. White balance 5000K for neutral color accuracy. No rolling shutter artifacts.
LIGHTING (studio control for reflections):
Two large softboxes at 45° left/right for smooth wrap; overhead 4×4 diffusion (silk) to create a soft highlight band on the glass; slim back kicker for a delicate rim on the jar shoulders; black negative fill on both sides to carve edge contrast; flags to prevent hotspots. 2:1 key-to-fill. High-CRI (95+) sources. Allow tasteful speculars and clean, realistic reflections—no blown highlights.
MATERIALS & SURFACE:
Physically accurate glass IOR and thickness; faint inner condensation above the fill line; subtle meniscus at the rim. Pesto shows suspended herbs and micro-bubbles with slight parallax as it rotates. Acrylic base produces a soft, fade-out reflection into the seamless background (no visible horizon).
MOOD & STYLE:
Premium product-photography aesthetic—clean, minimal, appetizing, color-true greens with a gentle, warm lift to enhance the olive oil while keeping whites neutral. No branding, no text overlays.
Spoiler: Die Bildgenerierung begeistert uns noch mehr
Die Videofähigkeiten sind schon beeindruckend, aber bei der Bildgenerierung spielt WAN 2.2 nochmal in einer anderen Liga. Die Qualität kommt sehr nah an professionelle Fotografie heran.
Besonders bei Portraits, Produktfotos und Materialdarstellungen sind die Ergebnisse sehr überzeugend. Haut sieht natürlich aus, Licht wirkt physikalisch korrekt, Texturen sind detailreich.
→ Unser nächster Artikel behandelt ausführlich die Bildgenerierung mit WAN 2.2. Dort schauen wir uns an, was genau möglich ist und wie die Prompts funktionieren.
Und jetzt? Lass deine kreativen Videos Wirklichkeit werden!
Probier Wan 2.2 einfach selbst aus – die Video-Generierung klappt auch unkompliziert über Plattformen wie Freepik, mit denen ich selbst schon gute Erfahrungen gemacht habe. Wenn du dabei Unterstützung brauchst oder Fragen hast, melde dich gerne bei uns. Wir helfen dir gerne weiter, damit du professionelle KI-Videos schnell und einfach erstellen kannst.
Bereit für den nächsten Schritt?
Die Tools entwickeln sich schnell, und jedes hat seine Besonderheiten.
Im KI Marketing Bootcamp gehen wir systematisch vor: Von der Strategie über die Tool-Auswahl bis zur konkreten Umsetzung. Du lernst nicht nur die Theorie, sondern arbeitest an echten Projekten – mit direktem Feedback und praktischen Workflows, die du sofort einsetzen kannst.
Was erwartet dich?
- Praxisorientierte Anleitungen: Lerne, wie du KI-Tools strategisch einsetzt und das Beste aus ihnen herausholst
- Erprobte Workflows: Vom Konzept bis zur Umsetzung – mit sofort anwendbaren Strategien
- Individuelle Begleitung: Kleine Gruppen und persönliche Betreuung bei deinen Projekten
Deine Vorteile:
- Learning by Doing: Entwickle eigene Kampagnen, die auf deine spezifischen Ziele zugeschnitten sind
- Praxiswissen: Nutze echte Beispiele und Erfolgsstrategien für deine eigenen Projekte
- 100% Online: Flexibel lernen, wann es in deinen Zeitplan passt
Für Unternehmen: Maßgeschneiderte Workshops für dein Marketing-Team.
Dein Expertenteam: Vroni Hackl und Georg Neumann – dein Expertenteam und deine Guides durch KI im Marketing.
