Bedste AI-videogenerator: 45 klip, scoret blindt

• Testet 1. november 2026
• 5 modeller
• 3 opgaver, 3 forsøg hver
• 45 klip, 0 fejlet
• Rådata medfølger

Afspil alle fem, uden lyd

Samme prompt til alle fem. Dette er forsøg 1 af 3 for hver model, så vi har ikke valgt de flatterende. Tryk på afspil på et hvilket som helst klip.

Kort svar: Hvis du vil have én model til alt, så brug Kling 4.0, for dens dårligste forsøg (3.6) slog alle andre modellers dårligste. Hvis prisen betyder mere, så brug Wan 3.0, som matchede Kling samlet set til $0.60 pr. klip. Seedance 2.5 ligger lige bag dem og er jævnbyrdig på talende klip. MiniMax H3 fører i scener med bevægeligt kamera. Oplysning: Hermoso sælger adgang til alle fem, og Seedance 2.5 er vores standard. Den vandt ikke en opgave alene.

Hvem bør bruge hvad

Én model til alt

Kling 4.0

Det jævneste resultat i testen. Dets laveste forsøg scorede 3.6, og ingen opgave var et svagt punkt. Det delte førstepladsen samlet med Wan 3.0.

Bedste værdi

Wan 3.0

Jævnbyrdig med Kling samlet set til $0.60 for 6 sekunder, testens skarpeste detaljer og den bedste naturlige lyd. MiniMax H3 er billigere pr. sekund, men scorede lavere uden for scener.

En person der taler foran kameraet

Uafgjort: Wan 3.0, Seedance 2.5, Kling 4.0

Alle tre sagde replikken ord for ord. Wan gav det største melstøv og de melede arme, vi bad om, med et smil fra en stock-reklame. Seedance lignede mest film, men rodede med tællingen i et af tre forsøg.

Produkter med småt tryk

Kling 4.0, Wan 3.0 tæt efter

Alle Kling-forsøg fik alle tre etiketlinjer rigtige, med en ægte hånd og et ægte push-in. Det fulgte også briefen bedst. Seedance lavede testens bedste honning, i et af tre forsøg.

Et bevægeligt kamera i en travl scene

MiniMax H3, derefter Wan og Kling

De højeste afsløringer, med rytteren stadig i billedet til sidst. Den havde også 8 sekunder, hvor de andre havde 6, hvilket hjælper et langsomt kameratræk. Wan 3.0 og Kling 4.0 ligger inden for vores uafgjort-grænse.

Når du har brug for det hurtigt

Veo 3.1

Blev færdig på cirka 66 sekunder i median, mod 156 til 238 for de øvrige. Det ser flot ud på et billede, men det sprang trin i briefen over oftere end nogen anden, til den højeste pris.

Scorerne

Hvert klip fik 1 til 5 på seks kriterier fra to adskilte blinde dommere. En opgavescore er gennemsnittet af dens tre forsøg, så et dårligt forsøg tæller med. Vi skrev reglen før nogen rendering: en forskel under 0.3 er uafgjort, og orange markerer topscoren og alt, der er uafgjort med den.

Model	Taler	Produkt	Scene	Samlet	Dårligste forsøg
Wan 3.0	4.17	3.67	3.72	3.85	3.29
Kling 4.0	3.95	3.86	3.66	3.82	3.62
Seedance 2.5	4.08	3.33	3.31	3.57	3.00
MiniMax H3	3.36	3.07	3.89	3.44	2.96
Veo 3.1	3.36	2.65	3.10	3.04	2.50

Samlet er det simple gennemsnit af de 45 scorer. Fjerner man lyden helt, overhaler Kling Wan lidt, mens de øvrige beholder deres placeringer: Kling 4.0 3.88, Wan 3.0 3.85, Seedance 2.5 3.65, MiniMax H3 3.52, Veo 3.1 3.05. Seedance 2.5 ligger inden for uafgjort-grænsen samlet set, og dens dårligste forsøg er grunden til, at vi placerer den bag de to førende.

Opgave 1: en person der taler og tæller på fingrene

En bager siger én præcis replik, viser én, to og så tre fingre, og klapper et brød, så melet hvirvler op. Den tester læbesynkronisering, ord, håndanatomi og små fysiske effekter i ét låst skud. Vinder: et tre-vejs uafgjort mellem Wan 3.0, Seedance 2.5 og Kling 4.0. Alle 15 klip sagde replikken ord for ord. Fingrene delte feltet: Veo 3.1 nåede aldrig tre i noget forsøg, og MiniMax H3 viste en fuld hånd i to af tre.

Forsøg 1Forsøg 2Forsøg 3

Afspil alle fem, uden lyd

Scorer for taler

1. Wan 3.04.17
2. Seedance 2.54.08
3. Kling 4.03.95
4. MiniMax H33.36
5. Veo 3.13.36

Gennemsnit af tre forsøg. Orange er førstepladsen eller inden for 0.3 af den.

Opgave 2: et honningglas med trykt etiket

Et langsomt push-in på et mærket glas, mens honning drypper fra en øse. Den tester småt skrift, om bogstaverne holder stille, og tyk væske. Etiketten er vores eget opdigtede produkt. Vinder: Kling 4.0, med Wan 3.0 inden for uafgjort-grænsen. Stavningen var tæt på løst: 14 af 15 klip fik alle tre linjer rigtige. Honningen var det ikke. Næsten intet foldede sig og rullede sig op, som bedt om, og Veo 3.1 kørte aldrig ind i noget forsøg.

Forsøg 1Forsøg 2Forsøg 3

Afspil alle fem, uden lyd

Scorer for produkt

1. Kling 4.03.86
2. Wan 3.03.67
3. Seedance 2.53.33
4. MiniMax H33.07
5. Veo 3.12.65

Gennemsnit af tre forsøg. Orange er førstepladsen eller inden for 0.3 af den.

Opgave 3: at følge en cyklist gennem et regnfuldt marked

Ét ubrudt skud bag en kurer i regnen: en sving uden om en vogn, en hund der krydser, ét ring fra en klokke, og så hæver kameraet sig over paraplyerne for at vise hele gaden. Vinder: MiniMax H3, med Wan 3.0 og Kling 4.0 inden for uafgjort-grænsen. Ingen fik alle momenterne med. Kling og H3 holdt sig tættest på briefen. Wan var den skarpeste, men byttede det almindelige lærred ud med farvede telte og opdigtede skilte. Seedance blev ved med at beskære rytteren, og Veo mistede rytteren helt i to af tre.

Forsøg 1Forsøg 2Forsøg 3

Afspil alle fem, uden lyd

Scorer for scene

1. MiniMax H33.89
2. Wan 3.03.72
3. Kling 4.03.66
4. Seedance 2.53.31
5. Veo 3.13.10

Gennemsnit af tre forsøg. Orange er førstepladsen eller inden for 0.3 af den.

Efter hvad vi scorede

Gennemsnit på tværs af alle 45 klip. Detalje adskiller modellerne mindst. At følge briefen adskiller dem mest, og det gap er det meste af grunden til, at rangordenen ser ud, som den gør.

Den dyreste model kom sidst

Veo 3.1 koster fire gange så meget pr. klip som Wan 3.0 og scorede lavest i hver opgave, på niveau med MiniMax H3 på talende klip. Billederne ser dejlige ud, med det mest troværdige mel i testen. Problemet er trinene: den talte aldrig til tre, kørte aldrig ind på honningen og mistede rytteren i regnen. Hvis dit skud er en enkel idé, kan det tjene dig bedre, end denne test antyder. Vores prompts var bevidst fulde af trin.

Stavning blev let, væsker blev det ikke

Vi regnede med, at etiketten ville være den svære del. Fjorten af 15 klip trykte alle tre linjer korrekt, og kun ét Veo-forsøg tilføjede et løst ord. Honning, der folder sig og ruller sig op, når den lander, er det, næsten alle missede. De fleste klip gav os en lige søjle eller en stiv form, der næsten ikke ændrede sig. Ét Seedance 2.5-forsøg fik det rigtigt, og det er også problemet: det var ét forsøg ud af tre.

Om musikken på nogle honningklip

Vi bad om køkkenlyde og ingen musik. På 8 af de 15 honningklip returnerede modellen næsten ingen lyd, under -40 dB, og vores renderingsforløb lagde et stille musikspor under den: alle tre MiniMax H3-forsøg, to Seedance 2.5, to Veo 3.1 og et Wan 3.0. Ingen af Kling-forsøgene havde brug for det. Musikken er vores, ikke modellens. Disse klip mistede point for lyd på grund af stilheden, og de er markeret i rådatatabellen. Tale- og scene-klip blev ikke berørt.

Sådan kørte vi den

Vi skrev de tre prompts og scoringsreglerne først, kørte ét pilotklip for at tjekke pipelinen og renderede derefter resten. Pilotklippet er et af de 45. Intet fejlede, så intet blev forsøgt igen eller droppet.

Alle klip bad om 16:9 i 720p og 6 sekunder, uden brandkontekst og uden omskrivning af prompten. To undtagelser: MiniMax H3 kørte i 8 sekunder, fordi den ikke havde en 6-sekunders mulighed, og den leverer 1344x768 i stedet for 1280x720. Wan 3.0 leverer med 30 billeder i sekundet, de øvrige med 24.

Hver opgave siger det samme til hver model, skrevet på den måde, som den pågældende models producent anbefaler, for en fair test lader ikke én model læse en anden models dialekt. Kling har ingen guide til 4.0, så vi brugte dens 3.0-vaner. Veo 3.1 er den fulde model, ikke dens hurtige version.

Vi sendte seeds 1101, 2202 og 3303 til Veo 3.1, Wan 3.0 og MiniMax H3. Vores renderingsforløb sender ikke en seed til Seedance 2.5 eller Kling 4.0, så disse forsøg er tre uafhængige kørsler. Ingen model bekræftede den seed, den brugte.

Scoringen var blind og blev udført af to adskilte AI-dommere, ikke af mennesker. Hvert klip blev reduceret til 12 billeder fra start til slut, et spektrogram af dets lyd, dets lydstyrke og et transskript, under en tilfældig kode. Ingen af dommerne så, hvilken model der havde lavet et klip, før deres scorer var på plads. De to dommere var tæt enige, med en korrelation på 0.91 på tværs af klip og et gennemsnitligt gap på 0.36 point pr. kriterie. Scorerne på denne side er deres gennemsnit, og hver dommers tal står i CSV-filen.

Nogle kontroller krævede ingen vurdering. Et lokalt transskript bekræftede den præcise replik i alle 15 talende klip. Scenedetektion fandt ingen klipninger i nogen af klippene, selvom ét Veo 3.1-forsøg afspilles som et hop, og det har vi markeret. Begge dommere talte etiketlinjerne hver for sig og var enige i alle 15.

Prompterne, ord for ord

Kopiér en hvilken som helst af dem. Hvert link åbner den prompt på den pågældende model i Hermoso.

A baker speaks to camera in a small bakery kitchen and counts three ingredients on her fingers. She is a woman in her late thirties with short curly dark hair, flour on her forearms and a navy linen apron.

Shot 1 (0-6 seconds): Locked-off medium shot at eye level, 35mm lens, shallow depth of field. She stands behind a floured wooden counter with a round crusty loaf in front of her and looks straight into the lens. She raises her right hand and shows one finger, then two fingers, then three fingers, one finger for each ingredient she names. Then she lowers both hands and pats the top of the loaf twice, and a small puff of flour rises into the window light.

Dialogue: The baker says in a warm, dry voice: "Just three things. Flour. Water. Salt. That's the whole loaf." Accurate lip sync.

Sound: her voice and the soft room tone of the kitchen.

Style: realistic documentary, warm morning window light, one continuous shot.

Negative prompt: subtitles, captions, on-screen text, watermark, background music, cuts.

Medium shot at eye level, locked-off camera, one continuous shot, realistic documentary look, warm morning window light in a small bakery kitchen. A woman in her late thirties with short curly dark hair, flour on her forearms and a navy linen apron stands behind a floured wooden counter with a round crusty loaf in front of her and looks straight into the lens. As she speaks she raises her right hand and shows one finger, then two fingers, then three fingers, one finger for each ingredient she names. Then she lowers both hands and pats the top of the loaf twice, and a small puff of flour rises into the window light.

Baker (warm, dry voice): "Just three things. Flour. Water. Salt. That's the whole loaf."

Only her voice and the soft room tone of the kitchen are heard.

A baker speaks to camera in a small bakery kitchen and counts three ingredients on her fingers, realistic documentary style, locked-off medium shot at eye level, warm morning window light.

0-1 seconds: She stands behind a floured wooden counter with a round crusty loaf in front of her and looks straight into the lens.
1-4 seconds: She raises her right hand and shows one finger, then two fingers, then three fingers, one finger for each ingredient she names.
4-6 seconds: She lowers both hands and pats the top of the loaf twice, and a small puff of flour rises into the window light.

Dialogue: The baker says in a warm, dry voice: "Just three things. Flour. Water. Salt. That's the whole loaf."

Constant throughout: she is a woman in her late thirties with short curly dark hair, flour on her forearms and a navy linen apron. One continuous shot, the camera stays still. Sound: her voice and the soft room tone of the kitchen. No subtitles. No background music.

integrated_multimodal_description: [Shot 1] Live-action, realistic documentary style, a Static Shot at eye level holds a medium shot of a small bakery kitchen in warm morning window light. A woman in her late thirties with short curly dark hair, flour on her forearms and a navy linen apron (S1), on screen, with a warm, dry voice, stands behind a floured wooden counter with a round crusty loaf in front of her. She looks straight into the lens and says: <d>[English] Just three things. Flour. Water. Salt. That's the whole loaf.</d> As she speaks she raises her right hand and shows one finger, then two fingers, then three fingers, one finger for each ingredient she names. Then she lowers both hands and pats the top of the loaf twice, and a small puff of flour rises into the window light.

overall_soundscape: The soft room tone of a small kitchen. Her palms pat lightly on the bread crust.

non_diegetic_music: N/A

A medium shot at eye level with a locked-off camera in one continuous take. In a small bakery kitchen lit by warm morning window light, a woman in her late thirties with short curly dark hair, flour on her forearms and a navy linen apron stands behind a floured wooden counter with a round crusty loaf in front of her and looks straight into the lens. She raises her right hand and shows one finger, then two fingers, then three fingers, one finger for each ingredient she names. She then lowers both hands and pats the top of the loaf twice, and a small puff of flour rises into the window light. Realistic documentary style, 35mm lens, shallow depth of field.

The baker says in a warm, dry voice: "Just three things. Flour. Water. Salt. That's the whole loaf."

Audio: her voice and the soft room tone of the kitchen, and nothing else.

A macro product shot of an open jar of honey with a printed label while a ribbon of honey falls into it.

Shot 1 (0-6 seconds): Slow push-in, 100mm macro lens. The frame shows an open, squat, clear glass jar of golden honey on a dark slate counter in soft daylight from the left. A cream paper label on the front of the jar is printed in black with exactly three lines of text: "HERMOSO" in large serif capitals, "WILDFLOWER HONEY" in smaller capitals below it, and "NET WT 12 OZ" in small capitals at the bottom. A hand at the top of the frame holds a wooden honey dipper above the jar, and a thin ribbon of thick honey falls into the open jar, where it folds and coils on the surface before slowly sinking. The camera pushes in slowly toward the label until the three lines of text fill the lower third of the frame, sharp and legible, with the letters fixed in place on the glass.

Sound: a quiet kitchen with a kettle simmering softly and birdsong through an open window.

Style: photoreal commercial, soft daylight, one continuous shot.

Negative prompt: subtitles, captions, watermark, background music, extra text on the label, cuts.

Macro product shot, one continuous shot, photoreal commercial look. The frame shows an open, squat, clear glass jar of golden honey on a dark slate counter in soft daylight from the left. A cream paper label on the front of the jar is printed in black with exactly three lines of text: "HERMOSO" in large serif capitals, "WILDFLOWER HONEY" in smaller capitals below it, and "NET WT 12 OZ" in small capitals at the bottom. A hand at the top of the frame holds a wooden honey dipper above the jar, and a thin ribbon of thick honey falls into the open jar, where it folds and coils on the surface before slowly sinking. The camera pushes in slowly toward the label until the three lines of text fill the lower third of the frame, sharp and legible, with the letters fixed in place on the glass.

Sound: a quiet kitchen with a kettle simmering softly and birdsong through an open window.

A macro product shot of an open jar of honey with a printed label while a ribbon of honey falls into it, photoreal commercial style, slow push-in, soft daylight.

0-2 seconds: The frame shows an open, squat, clear glass jar of golden honey on a dark slate counter in soft daylight from the left. A cream paper label on the front of the jar is printed in black with exactly three lines of text: "HERMOSO" in large serif capitals, "WILDFLOWER HONEY" in smaller capitals below it, and "NET WT 12 OZ" in small capitals at the bottom.
2-4 seconds: A hand at the top of the frame holds a wooden honey dipper above the jar, and a thin ribbon of thick honey falls into the open jar, where it folds and coils on the surface before slowly sinking.
4-6 seconds: The camera keeps pushing in slowly toward the label until the three lines of text fill the lower third of the frame, sharp and legible.

Constant throughout: the label's letters stay fixed in place on the glass. One continuous shot. Sound: a quiet kitchen with a kettle simmering softly and birdsong through an open window. No subtitles. No background music.

integrated_multimodal_description: [Shot 1] Live-action, photoreal commercial style, a macro shot shows an open, squat, clear glass jar of golden honey on a dark slate counter in soft daylight from the left. A cream paper label on the front of the jar is printed in black with exactly three lines of text: "HERMOSO" in large serif capitals, "WILDFLOWER HONEY" in smaller capitals below it, and "NET WT 12 OZ" in small capitals at the bottom. A hand at the top of the frame holds a wooden honey dipper above the jar, and a thin ribbon of thick honey falls into the open jar, where it folds and coils on the surface before slowly sinking. The camera performs a Push In at slow speed toward the label until the three lines of text fill the lower third of the frame, sharp and legible, with the letters fixed in place on the glass.

overall_soundscape: A quiet kitchen with a kettle simmering softly. Birdsong drifts through an open window.

non_diegetic_music: N/A

A macro product shot in one continuous take with a slow push-in, 100mm macro lens. The frame shows an open, squat, clear glass jar of golden honey on a dark slate counter in soft daylight from the left. A cream paper label on the front of the jar is printed in black with exactly three lines of text: "HERMOSO" in large serif capitals, "WILDFLOWER HONEY" in smaller capitals below it, and "NET WT 12 OZ" in small capitals at the bottom. A hand at the top of the frame holds a wooden honey dipper above the jar, and a thin ribbon of thick honey falls into the open jar, where it folds and coils on the surface before slowly sinking. The camera pushes in slowly toward the label until the three lines of text fill the lower third of the frame, sharp and legible, with the letters fixed in place on the glass. Photoreal commercial style.

Audio: a quiet kitchen with a kettle simmering softly and birdsong through an open window, and nothing else.

A low tracking shot follows a bicycle courier through a crowded rainy night market and rises above the crowd to reveal the whole street.

Shot 1 (0-6 seconds): Wide 24mm lens. In a narrow night-market street at dusk in heavy rain, lined with food stalls under plain canvas awnings and strings of warm lanterns, a bicycle courier in a yellow rain jacket rides away from the camera while the camera follows behind at low height at the same speed. The rider swerves left around a vendor pushing a cart, then a small dog runs across the street in front of the bicycle and the rider rings the bell once. Steam rises from a noodle stall and the wet cobblestones reflect the lanterns. In the last two seconds the camera rises smoothly above a crowd of people with umbrellas and reveals the whole street stretching ahead, rain streaks catching the lantern light.

Sound: heavy rain on canvas, one bicycle bell, sizzling woks, crowd chatter.

Style: cinematic realistic, rainy dusk, one continuous shot.

Negative prompt: subtitles, captions, on-screen text, watermark, background music, cuts.

One continuous tracking shot, cinematic realistic look. In a narrow night-market street at dusk in heavy rain, lined with food stalls under plain canvas awnings and strings of warm lanterns, a bicycle courier in a yellow rain jacket rides away from the camera while the camera follows behind at low height at the same speed. The rider swerves left around a vendor pushing a cart, then a small dog runs across the street in front of the bicycle and the rider rings the bell once. Steam rises from a noodle stall and the wet cobblestones reflect the lanterns. In the last two seconds the camera rises smoothly above a crowd of people with umbrellas and reveals the whole street stretching ahead, rain streaks catching the lantern light.

Sound: heavy rain on canvas, one bicycle bell, sizzling woks, crowd chatter.

A low tracking shot follows a bicycle courier through a crowded rainy night market and rises above the crowd to reveal the whole street, cinematic realistic style, one continuous shot.

0-2 seconds: In a narrow night-market street at dusk in heavy rain, lined with food stalls under plain canvas awnings and strings of warm lanterns, a bicycle courier in a yellow rain jacket rides away from the camera. The camera follows behind at low height at the same speed. The rider swerves left around a vendor pushing a cart.
2-4 seconds: Ahead, a small dog runs across the street in front of the bicycle and the rider rings the bell once. Steam rises from a noodle stall and the wet cobblestones reflect the lanterns.
4-6 seconds: The camera rises smoothly above a crowd of people with umbrellas and reveals the whole street stretching ahead, rain streaks catching the lantern light.

Constant throughout: the same rider, the same rainy dusk light. Sound: heavy rain on canvas, one bicycle bell, sizzling woks, crowd chatter. No subtitles. No background music.

integrated_multimodal_description: [Shot 1] Live-action, cinematic realistic style, a Tracking Shot at low height follows a bicycle courier in a yellow rain jacket who rides away from the camera through a narrow night-market street at dusk in heavy rain, lined with food stalls under plain canvas awnings and strings of warm lanterns, with the camera moving at the same speed. The rider swerves left around a vendor pushing a cart, then a small dog runs across the street in front of the bicycle and the rider rings the bell once. Steam rises from a noodle stall and the wet cobblestones reflect the lanterns. Near the end the camera performs a Pedestal Up with large amplitude at slow speed, rising above a crowd of people with umbrellas and revealing the whole street stretching ahead, rain streaks catching the lantern light.

overall_soundscape: Heavy rain drums on canvas and one bicycle bell rings. Woks sizzle under the crowd's chatter.

non_diegetic_music: N/A

One continuous tracking shot with a wide 24mm lens, cinematic realistic style. In a narrow night-market street at dusk in heavy rain, lined with food stalls under plain canvas awnings and strings of warm lanterns, a bicycle courier in a yellow rain jacket rides away from the camera while the camera follows behind at low height at the same speed. The rider swerves left around a vendor pushing a cart, then a small dog runs across the street in front of the bicycle and the rider rings the bell once. Steam rises from a noodle stall and the wet cobblestones reflect the lanterns. In the last two seconds the camera rises smoothly above a crowd of people with umbrellas and reveals the whole street stretching ahead, rain streaks catching the lantern light.

Audio: heavy rain on canvas, one bicycle bell, sizzling woks and crowd chatter, and nothing else.

Hvad denne test ikke kan fortælle dig

Tre prompts dækker ikke alt, og tre forsøg pr. opgave er en lille stikprøve, så læs et gap på 0.3 eller mindre som ingen forskel overhovedet. Begge dommere var AI-modeller, der læste billeder og målinger. Ingen har lyttet til klippene med øret: lyden blev vurderet ud fra transskripter, spektrogrammer og lydstyrke, så stemmekvalitet er det svageste punkt i vores scoring, og 12 billeder kan overse en fejl mellem dem. Vores mål for læbesynkronisering var groft, og vi stolede ikke på det. Det placerede Seedance 2.5 og Kling 4.0 tættest, og Wan 3.0 op til et halvt sekund ude af synk, så se klippene selv, hvis læbesynkronisering er vigtig for dig. Blindheden havde ét kendt hul: MiniMax H3's længde på 8 sekunder gjorde dens klip genkendelige, og de ekstra to sekunder kan have hjulpet den på den langsomme afsløring. Alt, der ikke står i tabellen, herunder Seedance 2.0 og HeyGen Video, blev ikke testet. Modeller og priser ændrer sig hurtigt, og disse resultater stammer fra én dag.