Best AI video generator: 45 clips, scored blind

• Tested 1 November 2026
• 5 models
• 3 jobs, 3 takes each
• 45 clips, 0 failed
• Raw data included

Play all five, muted

The same prompt on all five. This is take 1 of 3 for each model, so we did not pick the flattering ones. Press play on any clip.

The short answer: if you want one model for everything, use Kling 4.0, because its worst take (3.6) beat every other model's worst. If price matters more, use Wan 3.0, which matched Kling overall at $0.60 a clip. Seedance 2.5 sits just behind them and is level on talking clips. MiniMax H3 leads on moving-camera scenes. Disclosure: Hermoso sells access to all five, and Seedance 2.5 is our default. It did not win a job outright.

Who should use which

One model for everything

Kling 4.0

The steadiest result of the test. Its lowest take scored 3.6, and no job was a weak spot. It tied Wan 3.0 for first overall.

Best value

Wan 3.0

Level with Kling overall at $0.60 for 6 seconds, the sharpest detail of the test and the best natural sound. MiniMax H3 is cheaper per second but scored lower outside scenes.

A person talking on camera

Tie: Wan 3.0, Seedance 2.5, Kling 4.0

All three said the line word for word. Wan gave the biggest flour puff and the floury arms we asked for, with a stock-ad grin. Seedance looked the most like film but fluffed the count in one take of three.

Products with fine print

Kling 4.0, Wan 3.0 close

Every Kling take kept all three label lines right, with a real hand and a real push-in. It also matched the brief best. Seedance made the best honey of the test, in one take out of three.

A moving camera in a busy scene

MiniMax H3, then Wan and Kling

The tallest reveals, with the rider still in shot at the end. It also had 8 seconds where the others had 6, which helps a slow move. Wan 3.0 and Kling 4.0 are inside our tie line.

When you need it fast

Veo 3.1

Finished in about 66 seconds at the median, against 156 to 238 for the rest. It looks great in a still, but it skipped steps in the brief more often than anyone, at the highest price.

The scores

Each clip got 1 to 5 on six criteria from two separate blind judges. A job score is the average of its three takes, so a bad take counts. We wrote the rule before any render: a gap under 0.3 is a tie, and orange marks the top score and anything tied with it.

Model	Talking	Product	Scene	Overall	Worst take
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

Overall is the plain average of the 45 scores. Take audio out completely and Kling edges past Wan while the rest keep their places: 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 is inside the tie line overall, and its worst take is why we rank it behind the two leaders.

Job 1: a person talking and counting on her fingers

A baker says one exact line, shows one, two, then three fingers, and pats a loaf so that flour rises. It tests lip sync, words, hand anatomy and small physics in one locked shot. Winner: a three-way tie between Wan 3.0, Seedance 2.5 and Kling 4.0. All 15 clips said the line word for word. Fingers split the field: Veo 3.1 never reached three in any take, and MiniMax H3 opened to a full hand in two of three.

Take 1Take 2Take 3

Play all five, muted

Talking scores

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

Average of three takes. Orange is first place or within 0.3 of it.

Job 2: a honey jar with a printed label

A slow push-in on a labelled jar while honey falls from a dipper. It tests small text, whether letters hold still, and thick liquid. The label is our own made-up product. Winner: Kling 4.0, with Wan 3.0 inside the tie line. Spelling was close to solved: 14 of 15 clips got all three lines right. Honey was not. Almost nothing folded and coiled as asked, and Veo 3.1 never pushed in on any take.

Take 1Take 2Take 3

Play all five, muted

Product scores

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

Average of three takes. Orange is first place or within 0.3 of it.

Job 3: following a cyclist through a rainy market

One unbroken shot behind a courier in the rain: a swerve around a cart, a dog crossing, one ring of a bell, then the camera rises over the umbrellas to show the whole street. Winner: MiniMax H3, with Wan 3.0 and Kling 4.0 inside the tie line. Nobody got every beat. Kling and H3 kept closest to the brief. Wan was the sharpest but swapped plain canvas for coloured tents and made-up signs. Seedance kept cropping the rider, and Veo lost the rider altogether in two of three.

Take 1Take 2Take 3

Play all five, muted

Scene scores

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

Average of three takes. Orange is first place or within 0.3 of it.

By what we scored

Averages across all 45 clips. Detail separates these models the least. Following the brief separates them the most, and that gap is most of why the ranking looks the way it does.

The most expensive model came last

Veo 3.1 costs four times as much per clip as Wan 3.0 and scored lowest on every job, level with MiniMax H3 on talking. The stills look lovely, with the most believable flour in the test. The trouble is steps: it never counted to three, never pushed in on the honey, and lost the rider in the rain. If your shot is one simple idea, it may serve you better than this test suggests. Our prompts were deliberately full of steps.

Spelling got easy, liquids did not

We expected the label to be the hard part. Fourteen of 15 clips printed all three lines correctly, and only one Veo take added a stray word. Honey that folds and coils as it lands is what nearly everyone missed. Most clips gave us a straight column, or a stiff shape that barely changed. One Seedance 2.5 take got it right, which is also the problem: it was one take of three.

About the music on some honey clips

We asked for kitchen sounds and no music. On 8 of the 15 honey clips the model returned next to no sound, under -40 dB, and our render path added a quiet music track under it: all three MiniMax H3 takes, two Seedance 2.5, two Veo 3.1 and one Wan 3.0. None of the Kling takes needed it. That music is ours, not the model's. Those clips lost points on audio for the silence, and they are marked in the raw table. Talking and scene clips were not touched.

How we ran it

We wrote the three prompts and the scoring rules first, ran a single pilot clip to check the pipeline, then rendered the rest. The pilot is one of the 45. Nothing failed, so nothing was retried or dropped.

Every clip asked for 16:9 at 720p and 6 seconds, with no brand context and no rewriting of the prompt. Two exceptions: MiniMax H3 ran at 8 seconds because it had no 6-second option, and it delivers 1344x768 rather than 1280x720. Wan 3.0 delivers at 30 frames a second, the rest at 24.

Each job says the same thing to every model, written the way that model's maker recommends, because a fair test does not make one model read another's dialect. Kling has no guide for 4.0, so we used its 3.0 habits. Veo 3.1 is the full model, not its fast version.

We sent seeds 1101, 2202 and 3303 to Veo 3.1, Wan 3.0 and MiniMax H3. Our render path does not pass a seed to Seedance 2.5 or Kling 4.0, so those takes are three independent runs. No model confirmed the seed it used.

Scoring was blind and done by two separate AI judges, not people. Each clip was reduced to 12 frames from start to end, a spectrogram of its sound, its loudness and a transcript, under a random code. Neither judge saw which model made a clip until its scores were in. The two judges agreed closely, with a correlation of 0.91 across clips and an average gap of 0.36 points per criterion. Scores on this page are their average, and each judge's number is in the CSV.

Some checks did not need judgment. A local transcript confirmed the exact line on all 15 talking clips. Scene detection found no cuts in any clip, though one Veo 3.1 take plays as a jump and we marked it. Both judges counted label lines separately and matched on all 15.

The prompts, word for word

Copy any of these. Each link opens that prompt on that model in 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.

What this test can't tell you

Three prompts do not cover everything, and three takes per job is a small sample, so read a gap of 0.3 or less as no gap at all. Both judges were AI models reading frames and measurements. Nobody listened to these clips by ear: sound was judged from transcripts, spectrograms and loudness, so voice quality is the weakest part of our scoring, and 12 frames can miss a glitch between them. Our lip sync measure was rough and we did not rely on it. It put Seedance 2.5 and Kling 4.0 closest, and Wan 3.0 up to half a second out, so watch the clips if lip sync matters to you. Blinding had one known hole: MiniMax H3's 8-second length made its clips recognisable, and those extra two seconds may have helped it on the slow reveal. Anything not in the table, including Seedance 2.0 and HeyGen Video, was not tested. Models and prices change quickly, and these results are from one day.