Google Veo 3 video generation measurement, so powerful!

At the 2025 Google I/O developer conference, Google unveiled Veo 3 – a remarkable video generation model. It not only has strong physical understanding capabilities and can generate smooth and realistic animations, but also supports native audio generation, including ambient sounds, sound effects and even character dialogue, which greatly enhances the immersion and realism of the video. This article will demonstrate Veo 3’s powerful capabilities in video generation through a series of measured cases, explore its performance in different scenarios, and analyze its advantages and disadvantages in detail processing, complex prompt understanding, and grand scene control.

At the 2025 Google I/O developer conference a few days ago, Google released a series of advanced image and video generation tools, there are so many fun that I haven’t had time to experience them one by one, today I will try the recently super popular Veo 3 video generation. Try Imagen 4 and the Flow platform to share with you in the future. Let me introduce it briefly.

  • Veo 3 is Google’s latest video generation model, and officials say it has stronger physical understanding capabilities and generates smoother and more realistic animations.
  • Native Veo 3 already supports direct audio generation, including ambient sounds, sound effects, and even character dialogue, which can make AI-generated videos more immersive and realistic.
  • The model is open to AI Pro and AI Ultra subscribers.

The platform I use is Gemini, but it currently only supports Wensheng Graph, link: https://gemini.google.com/

And Flow can also support graph generation, first and last frames, link: https://labs.google/fx/tools/flow

Think about weekends (and I do want to be lazy!). Most of today’s tests are Wensheng videos. This issue is very fun, I hope you like it~!

Use on the Gemini platform:

谷歌Veo 3视频生成实测,8秒玩出新乐趣

Using Wensheng Video on the Flow platform:

Click to select [Wensheng Video], and then enter the prompt.

  • Vincent Videos: Generate videos with descriptive text prompts.
  • Image Raw Video: Supports first, last frame, or first and last frame references to generate dynamic content (250525 currently supports external image upload).
  • Element combination to generate videos: Extract the content and style of multiple images and combine them with prompts to generate videos.
谷歌Veo 3视频生成实测,8秒玩出新乐趣

Note this setting:

谷歌Veo 3视频生成实测,8秒玩出新乐趣

 

In Flow, you can also export to GIF format or 720P at the end of export, and 1080P needs to be exported after clicking Super Resolution. Flow also has ways to extend video online editing and other gameplay, so I’ll share it next time!

How can product managers do a good job in B-end digitalization?
All walks of life have taken advantage of the ride-hail of digital transformation and achieved the rapid development of the industry. Since B-end products are products that provide services for enterprises, how should enterprises ride the digital ride?

View details >

Here are 10 videos below, all generated using Veo 3. The video topics are as follows:

Final applause

Lost connections

Kitten astronaut

She doesn’t remember me

Reflection of time

Night shift fox

To-do list

A giant beast in the sea

Alien workers

A galloping koala

Before the video starts, I would like to talk about my general ideas. Because it is only 8 seconds, and Wensheng Video makes it more unknown, there is no way to control its overall style and main body through the picture at the beginning, so this has a certain accidental nature of card drawing, and it is easy to collapse, so my idea is:

1. Give as much content and restrictions as possible in your prompts. Prompts include, but are not limited to, visual style, story overview, and try to include the most advanced voiceover and subtitle descriptions that are currently available.

2. 8 seconds is very short, but you can also do some changes, because it is just a Wensheng video is not easy to continue, I hope that these 8 seconds can quickly convey a certain feeling, try to break the 8 seconds into 4 paragraphs in the prompt, and there is a scene change, emotional progression or turning point every two seconds.

Note that these prompts are not completely achievable, this is just my idealized situation, the prompt will write the content within 8 seconds, and the actual implementation can reach 70%-80%, which is already good.

Final applause

Story Title: “The Final Applause” Visual Style: Black-and-white sketch style Rough lines outline the classical theater and the aging actor, while the robotic audience is depicted with clean geometric shapes. The stark black-and-white contrast creates a tension of coldness and solitude.

Story Overview: Seconds 1-2: The aging actor stands at the center of the theater, a spotlight illuminating his gaunt face, while the surrounding audience seats are completely empty. Seconds 3-4: The camera shifts to the audience seats, revealing rows of robots sitting neatly, expressionlessly analyzing data. Seconds 5-6: The actor bows for the curtain call, the lights flicker, and the robots rise in unison to applaud, their mechanical clapping sounding like raindrops striking metal. Seconds 7-8: The actor closes his eyes, smiling with tears streaming down, as the camera pans from above to reveal the grand yet hollow theater.

Story Title: “The Last Applause” Visual Style: Black and white sketch style, with rough lines outlining classical theater and aging actors, while mechanized audiences are depicted in clean geometric shapes. The stark contrast of black and white creates a sense of indifference and loneliness.

Story Overview: Seconds 1-2: The aging actor stands in the center of the theater, spotlights illuminate his emaciated face, and the surrounding audience is completely empty. Seconds 3-4: The camera turns to the audience, showing a robot sitting neatly, analyzing the data expressionlessly. Seconds 5-6: The actors bow and call the curtain, the lights flicker, the robots give a standing ovation in unison, and the mechanical applause sounds like raindrops hitting metal. Seconds 7-8: The actor closes his eyes, smiling and tears flow, and the camera moves from above, revealing a magnificent but empty theater.

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/YJc5ByZuDBwK448LL6ne.mp4?_=2100:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

Lost connections

In the pitch-black pixel city, a lonely finger points at the flickering “CONNECT” button and gently presses it. A pink data beam carries anticipation across the city, illuminating countless empty windows. On the other end, the silhouette of a waiting pixel girl is lit by the beam. Suddenly, the screen glitches, color blocks tear apart, the signal bar plummets to zero, and the world falls silent. In the darkness, a real heartbeat echoes, scattered garbled codes converge spontaneously with the rhythm, forming two small hearts gazing at each other from afar. The heartbeat stops, the pixel hearts fade, and the screen goes completely black—perhaps the connection was already successful, simply because we are still searching for each other.

In the pitch-black pixel city, a lonely person points to the flashing “Connect” button and presses it gently. The pink data beam carries anticipation through the city, illuminating countless empty windows. At the other end, the silhouette of the waiting pixel girl is illuminated by a beam of light. Suddenly, the screen malfunctions, the color block tears, the signal bar plummets to zero, and the world falls silent. In the darkness, real heartbeats echo, and scattered garbled codes spontaneously converge with rhythm, forming two small hearts staring at each other in the distance. The heartbeat stops, the pixel heart fades, the screen goes completely dark – perhaps the connection has been made long ago, just because we’re still looking for each other.

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/6UU6HYi1qyEBOIlITJu5.mp4?_=2200:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

Kitten astronaut

“The Last Rescue” Seconds 1-2: Inside a dark wormhole, flickering lights illuminate the scene. A cat in a spacesuit floats inside the control cabin, its two paws furiously typing on the keyboard (soft paw pads tapping rapidly). Its little face is intensely focused, with the light reflecting off its helmet twisting like a nebula. Seconds 3-4: The system blares: “Wormhole collapsing! Navigation failed!” It lets out a sharp meow, flips around, and kicks the engine start button with a paw. Seconds 5-6: The ship begins to spin. The cat clings tightly to the edge of the screen, its fur bristling, eyes wide open. The camera zooms in as it declares, “I can’t give up… the galaxy still needs cats.” Seconds 7-8: A beam of white light envelops the entire spaceship. In the final frame, a photo appears inside its helmet: the cat basking in the sun with its owner. Subtitles emerge: “For home, chasing the last ray of light.”

“The Last Salvation” 1-2 seconds: Inside a dark wormhole, flashing lights illuminate the scene. A cat in a spacesuit floats in the control cabin, its claws tapping rapidly on the keyboard (soft paw pads hitting quickly). Its small face is focused, and the light on its helmet reflects a nebula-like twisted brilliance. 3-4 seconds: The system sounds an alarm: “The wormhole collapses! Navigation failed! It made a high-pitched meow, flipped over, and slammed the engine start button with its paws. Seconds 5-6: The ship starts spinning. The cat clings to the edge of the screen, its hair stands on end, and its eyes wide open. The camera zoomed in and it declared, “I can’t give up…… The galaxy also needs cats. Seconds 7-8: A beam of white light envelops the entire spacecraft. In the last frame, a photo appears inside the helmet: the cat enjoys the sun with its owner. The caption appears: “For home, chase the last ray of light.” ”

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/nr1yYiU2n5eekdmaAZod.mp4?_=2300:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

She doesn’t remember me

Story Title: “She Doesn’t Remember Me” Visual Style: Retro Cyberpunk Under the neon-lit rainy streets, an 80s CRT-style interface flickers. Characters wear old-fashioned metallic implants, complemented by a grainy film texture and red-blue halos. Story Overview: Seconds 1-2: A man walks into an abandoned memory restoration shop on a rainy neon night, holding a chip in his hand, his face weary. Seconds 3-4: A woman’s image appears on the screen, her face familiar yet devoid of emotion. He softly calls her name. Seconds 5-6: She looks at him, blinks, and coldly says, “User identification failed.” Seconds 7-8: He inserts the chip into his neck, the screen abruptly goes dark, and as the sound of rain echoes, he vanishes into the street’s interplay of light and shadow. Key Line: “She doesn’t remember me.”

Story Title: “She Doesn’t Remember Me” Visual Style: Retro Cyberpunk 80s CRT-style interface flashes on a neon-lit rainy night street. The character wears vintage metal implants with a grainy film texture and a red and blue light ring.

Story Overview: Seconds 1-2: A man walks into an abandoned memory recovery shop under the neon lights of a rainy night, holding a chip in his hand and looking tired. 3-4 seconds: An image of a woman appears on the screen, her face familiar but expressionless. He called out her name softly. Seconds 5-6: She looked at him, blinked, and said coldly, “User authentication failed.” “7-8 seconds: He inserts the chip into his neck, the screen suddenly goes black, and as the sound of rain echoes, he disappears into the interplay of light and shadow in the streets.

Key line: “She doesn’t remember me.” ”

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/Ia0xiAQEwaPnwCy1j4Wj.mp4?_=2400:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

Reflection of time

The camera focuses on a massive Time Mirror, reflecting the protagonist’s youthful laughter. As her fingertips lightly touch the surface, the reflection instantly ages, the face marked with traces of time. The mirror slowly cracks, and time flows out like liquid through the fissures, seeping into reality. Eventually, the mirror shatters into a black-and-white image. The protagonist stands quietly, as the screen gradually displays the words, “Memory is the reflection of time” .

The camera focuses on a giant mirror of time, reflecting the protagonist’s youthful laughter. When her fingertips lightly touch the mirror, the reflection ages instantly, and her face bears traces of time. The mirror slowly cracked, and time flowed out of the crack like a liquid, seeping into reality. Eventually, the mirror shattered into black and white images. The protagonist stands quietly, and the screen gradually shows the phrase “memory is a reflection of time”.

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/S83BKNqMmk9PGNX10FtT.mp4?_=2500:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

Night shift fox

“Night Shift Fox” Visual Style: Futuristic Neon Aesthetic + 80s Retro Tech Vibes The city at night is interwoven with purple, blue, and red lights, with reflective, glimmering streets. The fox wears a tailored suit, its tail sweeping light trails across the ground. The overall scene feels sci-fi yet detailed and realistic, with cold, striking colors and a composition full of tension.

Story Summary: Seconds 1-2: The camera tilts down from an overpass, showing a fox carrying a lunchbox walking along an empty street. Behind it, neon advertisements flash wildly with the slogan “Efficiency Above All.” Seconds 3-4: The fox sits on a street corner electrical box eating, surrounded by AI courier rabbits and robotic security dogs running past, with no one stopping to notice it. Seconds 5-6: It takes a bite of its sandwich, oil glistening at the corner of its mouth, then looks up at the virtual moon, pausing in silence. Seconds 7-8: It murmurs, “The city doesn’t sleep, so neither can I.” The lights reflect in its eyes, faintly bright and slightly wet.

Key Line: “The city doesn’t sleep, so neither can I.”

“Night Fox” Visual Style: Futuristic Neon Aesthetic + 80s Retro Tech Atmosphere The city at night is intertwined with purple, blue and red lights, reflecting the twinkling streets. The fox wore a tailored suit and its tail swept out a light trail on the ground. The overall scene feels sci-fi and detailed and realistic, with cold and vivid colors, and the composition is full of tension.

Story Summary: Seconds 1-2: The camera tilts down from the viaduct, showing a fox with a lunch box walking down an empty street. Behind it, neon advertisements flashed wildly, with the slogan “Efficiency First”. 3-4 seconds: The fox is sitting on the electric box on the street corner eating, surrounded by AI courier rabbits and robot police dogs running by, and no one stops to pay attention to it. Seconds 5-6: It takes a bite of a sandwich, the corners of its mouth shine with oil, then looks up at the virtual moon and is silent for a moment. Seconds 7-8: It whispers, “The city doesn’t sleep, and I can’t sleep.” “The light reflects in its eyes, slightly bright and slightly damp.

Key line: “The city doesn’t sleep, and I can’t sleep.” ”

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/jnL1EkCKP6KKzP8mZydn.mp4?_=2600:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

To-do list

“To-Do List” Visual Style: Paper Craft Animation Style All elements appear as if crafted from real handmade paper, cut, folded, and collaged: characters are silhouette collages, the task list is a tearable sticky note, and the background uses textured paper to create a timeline, alarm clock, and calendar imagery. The camera slowly zooms in, with each frame resembling a framed artwork.

Story Overview: Seconds 1-2: A “Today’s To-Do” list made of paper pieces gently falls onto the desk, densely packed with tasks like “Reply to Emails,” “Meeting Recap,” and “Health Check-In.” Seconds 3-4: A paper silhouette character (the protagonist) busily moves around, tearing off one task at a time from the list with increasing speed as tasks are completed. Seconds 5-6: The last sticky note reads “Breathe.” The paper figure pauses, hesitating as they look at it. Seconds 7-8: They gently tear off the “Breathe” note but instead of placing it in the completed tasks pile, they stick it to their chest and close their eyes. The entire screen freezes into a textured cover card.

Key Phrases (Text on Paper): Second 6: “Breathe” Second 8: “This, too, is worth completing.” (Appears embossed on the cover)

To-Do List Visual Style: Paper Art Animation Style All elements look as if they were made from real handmade paper, cut, folded and collaged: characters are silhouette collages, task lists are tearable sticky notes, and backgrounds are created with textured paper to create timelines, alarm clocks, and calendar images. The camera slowly zooms in, and each frame looks like a framed work of art.

Story Overview: Seconds 1-2: A “Today’s To-Do” list made of pieces of paper gently falls on the table, densely packed with tasks such as “Replying to Mail”, “Meeting Summary” and “Health Check”. Seconds 3-4: A paper silhouette character (the main character) is busy moving around, tearing a quest off the list faster and faster as the task is completed. Seconds 5-6: The last note says “breathe”. The paper character stops and looks at it hesitantly. 7-8 seconds: They gently tear off the “breathing” sticky note, but instead of putting it in the pile of completed tasks, stick it to their chest and close their eyes. The whole picture freezes into a textured cover card.

Key phrases (words on paper): 6th second: “Breathe” 8th second: “It’s worth accomplishing too.” (Appears on the cover in embossed form)

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/EaIgKo4z2f57n3RE0Dsq.mp4?_=2700:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

A giant beast in the sea

“Sea Monster” Visual Style: Hyper-realistic CG + Low-angle handheld perspective + Strong backlit composition The camera remains in a low-angle shot throughout, with extended focal length to emphasize the “endless height of the monster,” akin to the “divine fear” depicted in works like *Godzilla*, *Snowpiercer*, and *The Mountain Giant*. Story Synopsis: Seconds 1–2: The torrential rain has just stopped, the night sky looms heavily. A crew member looks up at the distant horizon; the sea surface seems to bulge upward. The shot slowly tilts up from behind him—something begins to rise from the water. Seconds 3–4: The sea monster fully stands up from the ocean. Its fin bones, rock-like armor plates, flickering deep-sea luminescent spots, and partially translucent biological tissues are revealed under the moonlight. Seconds 5–6: The camera pulls back into a low-angle wide shot. The crew member appears as small as a sesame seed. He gasps and stumbles backward, muttering, “It stood up… It really stood up…” His voice begins to crack. Seconds 7–8: The monster’s head finally emerges from the sea, its massive form nearly blotting out the sky. A partially folded wing unfurls, stirring up waves. The camera shakes violently amidst water vapor and glowing specks. The screen cuts to black just after a split-second overexposed flash, accompanied by the crew member’s scream. Key Dialogue (whispering in terror): “It stood up… It really stood up…” Cinematography: Opening with a low-angle wide shot → Slowly pushing in closer → Mid-section shifts to a upward view of the full body → Ending with intense shaking + overexposed white flash + black screen

“Beasts of the Sea” visual style: surreal CG + low-angle handheld view + strong backlit composition

The camera always shoots at low angles, using a long focal length to emphasize the “endless height of the monster”, similar to the “sacred fear” depicted in works such as Godzilla, Snowpiercer, and Mountain Giant.

Story overview: Seconds 1-2: The downpour has just stopped, and the night sky looks heavy. A crew member looks up at the distant horizon; The sea surface appears to be rising upwards. The camera slowly tilts upwards from behind him – something begins to rise from the water.

Seconds 3-4: The sea monster rises completely from the ocean. Its fin bones, rock-like armor, shimmering deep-sea glowing spots, and partially transparent biological tissue are revealed in the moonlight.

Sec. 5-6: The camera pulls back to a low-angle wide lens. The crew looked as small as sesame seeds. He gasped and stepped back, muttering, “It stands up…… It really stood up…… his voice began to tremble.

Seconds 7-8: The monster’s head finally emerges from the sea, and its massive body almost covers the sky. A partially folded wing spreads out and makes waves. The camera shakes violently in water vapor and luminous particles. The picture switches to a black screen after a flash of instant exposure, accompanied by the screams of the crew.

Key Dialogue (Whisper of Fear): “It stands up…… It really stood up……

Photography: Opening with a low-angle wide lens → slowly pushing closer to → Mid-section switching to a top-down full-body view → Ending with violent shaking + overexposed white light + black screen.

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/DaQ7qsagFZecSMu8Pcvg.mp4?_=2800:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

Alien workers

“Alien Worker” Visual Style: Handheld interview-style camera + defocused zoom effect + cartoonish anthropomorphic aliens (soft, round heads with big eyes). Natural street lighting, aliens resembling a blend of Pixar and POP MART designs, dressed in Earth delivery uniforms or carrying food delivery backpacks. The overall visuals are realistic, but the characters feel “deliberately out of place.” Story Summary (8-second structure): Seconds 1-2: The camera focuses on the street. A young reporter asks, “Which planet are you from?” The frame shakes slightly as the camera pans to the alien. Seconds 3-4: The alien, visibly exhausted and holding a drink bag, responds, “Sg’r’bl… On our planet, we only work 2 hours a day.” Seconds 5-6: Drooping its antennae, the alien sighs, “I just planned to do a temp job on Earth… but now rent, utilities, social security… it’s too much.” Seconds 7-8: The alien gazes at the sky, mumbling softly, “I wanna go home… I miss my mom’s plasma soup…” Background text appears: “Even aliens struggle with labor.” Key Lines (adorable anthropomorphic tone): “Sg’r’bl… On our planet, we only work 2 hours a day.”

Visual style: Handheld interview-style camera + defocus zoom effect + cartoonish humanoid alien (soft, rounded head and big eyes). Natural street light, aliens look like a combination of Pixar and POP MART designs, wearing Earth’s courier uniform or carrying a takeaway backpack. The overall visuals are realistic, but the characters feel “deliberately out of place”.

Story summary (8-second structure): 1-2 seconds: The camera focuses on the street. A young journalist asked, “What planet are you from?” The camera shakes slightly and the camera turns to the alien. 3-4 seconds: The alien is visibly tired, holding a drink bag in his hand, and replies: “Sg’r’bl…… On our planet, we only work 2 hours a day. Seconds 5-6: The alien lowers his tentacles and sighs, “I was going to be a temporary worker on Earth…… But now rent, utilities, social security…… Too much. 7-8 seconds: The alien stares at the sky and mutters softly: “I want to go home…… I miss my mother’s plasma soup ……” appears in the background: “Even aliens are struggling for labor.” ”

Key line (cute anthropomorphic tone): “Sg’r’bl…… On our planet, we only work 2 hours a day. ”

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/yYue7C5tqsY90e8Yv91I.mp4?_=2900:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

A galloping koala

“Running Koala” Visual Style: Hyper-realistic 3D + Handheld Cinematic Feel

The visuals feature intense camera shakes, rapid focus shifts, and a backdrop of a volcanic eruption with interwoven red, black, and gray tones. The koala’s fur is damp and muddy, its eyes filled with fear and struggle. The overall color palette is dominated by fiery orange-red glows and deep gray ash, reminiscent of “The Revenant” + “Dante’s Peak.”

Story Overview: Seconds 1-2: Amid shaky camera movements, the focus locks on a koala sprinting along the edge of a volcanic forest. Behind it, the ground cracks open, spewing magma, with the erupting volcano looming in the distance. Seconds 3-4: The camera rapidly zooms in for a close-up of the koala’s face—it glances back, its eyes a mix of terror and defiance. Ash drifts down from the sky as the camera shifts and blurs with its frantic movement. Seconds 5-6: Suddenly, from a low-angle shot, a burning tree trunk crashes down behind the koala, the firelight illuminating its silhouette in red. Seconds 7-8: The koala bursts out of the forest edge, leaping toward the camera. The screen goes black upon impact, and text appears: “Will you run toward hope, or into the flames?”

Key Dialogue (Subtitle): “Will you run toward hope, or into the flames?”

“Running Koala” Visual style: surreal 3D + handheld cinematic feel

Visuals: Includes intense camera shakes, quick focus switches, and a background of a volcanic eruption intertwined with shades of red, black, and gray. The koala’s fur is wet and muddy, and its eyes are full of fear and struggle. The overall color palette is dominated by fiery orange-red glow and dark gray volcanic ash, reminiscent of “The Revenant” and “Dante’s Peak.”

Story Overview: Seconds 1-2: In the shaky camera movement, the focus is locked on a koala running along the edge of a volcanic forest. Behind it, the ground cracked, magma spewed out, and volcanoes were erupting in the distance. Seconds 3-4: The camera zooms in quickly to a close-up of the koala’s face – it glances back, its eyes filled with a mixture of fear and defiance. Ash falls from the sky, and the camera blurs as it moves like crazy. Seconds 5-6: Suddenly, shooting from a low angle, a burning tree trunk falls behind the koala, and the flames illuminate its outline in red. Seconds 7-8: Koala rushes out of the edge of the forest and jumps towards the camera. On impact, the picture goes black and the text appears: “Will you run to hope, or into the flames?” ”

Key dialogue (subtitles): “Will you run towards hope or into the flames?” ”

Video player

Media error: Format(s) not supported or source(s) not found

Download file: https://www.woshipm.com/wp-content/uploads/2025/05/1MG95xFhNRU5kiKIhMOL.mp4?_=3000:0000:0000:00Use the up/down arrow keys to increase or decrease the volume.

brief summary

After trying this set of Veo 3 Vincent videos, I feel that it is really great. First of all, in terms of video quality, the picture quality is clear, and the general physics and motion simulation are very natural and smooth; At the same time, because of audio generation and lip synchronization, the realism of the video is greatly enhanced, which is very immersive; It also has a strong understanding of complex prompts, and can handle scene switching better. In short, it has reached the next level in terms of Wensheng video alone, and I look forward to other AI video generation tools following up as soon as possible (knocking down the price).

Of course, there are still many problems at present.

1. Detail issues. Occasional strange sound effects; After more than one object, the voice does not match the character; and the problem of physical dynamic simulation still exists, such as molding, transformation, as well as complex body movements and facial micro-expressions, emotional expressions, etc., there is still room for improvement in the realism of dynamics, and the direction is often unclear.

2. There is a bias in the understanding of complex prompts, and when there are many prompts and the storyboard is switched, the results may not match the requirements of the prompts.

3. The ability to control the details of relatively grand scenes needs to be improved.

Okay, today’s video is shared here, which one do you like the most? Looking forward to the exchange and sharing in the comment area~

End of text
 0