Video wird geladen...

Video konnte nicht geladen werden

Zur Startseite

⚡AI Video Model Comparison: Text to Video 9 Models included: • Google Veo 2 • Adobe Firefly Video • Hunyuan Video • Pika 2.1 Turbo • Genmo Mochi 1 • Runway Gen-3 • Kling 1.6 • Luma Ray2 •.Hailuo T2V-01 Director I used the same prompt on each model...

HBCoop_'s profile picture

Heather Cooper

54,606 subscribers

80,211 Aufrufe • vor 1 Jahr •via X (Twitter)

Wissenschaft & Technologie
LIVE

Anya Rossi• Live Now

Private livecam show

12 Kommentare

Profilbild von Loop'n'Mix
Loop'n'Mixvor 1 Jahr

Great test Heather! On this one Veo, Kling and Runwayml stood out the most for me 👍

Profilbild von Heather Cooper
Heather Coopervor 1 Jahr

Veo is super clean!

Profilbild von InterSystems Developers
InterSystems Developersvor 1 Jahr

📺 Watch this #video to learn how #OpenSource vs. proprietary #AI models compare one to another. We explore how proprietary models dominate leaderboards like Chatbot Arena, but open-source models are quickly catching up. Plus, we discuss how businesses can start with proprietary models and transition to open-source alternatives over time 👇 Fit AI into your development strategy!

Profilbild von Rupp 🐶
Rupp 🐶vor 1 Jahr

Love these comparison videos. Which have you seen consistently work the best? Kliing/MiniMax/Luma still?

Profilbild von Heather Cooper
Heather Coopervor 1 Jahr

Thank you! Kling has been solid for several months

Profilbild von Jett Collective
Jett Collectivevor 1 Jahr

Always appreciate these, Heather!

Profilbild von Heather Cooper
Heather Coopervor 1 Jahr

No problem Jett. I find it helpful to always have a side by side. Helps see where each tool has it's strengths and which is best used for a certain purpose.

Profilbild von Azer Koçulu
Azer Koçuluvor 1 Jahr

Great work Heather! Have you tested image to video as well?

Profilbild von Heather Cooper
Heather Coopervor 1 Jahr

Yes, but not for this prompt. I created it as a text to video prompt. I have many image to video comparisons though. I'll post one tomorrow

Profilbild von thirushadow
thirushadowvor 1 Jahr

Thank you very much 🙏

Profilbild von Andrew Holdun
Andrew Holdunvor 1 Jahr

Great comparisons. Any place to try out Veo 2? Been requesting for a long time and no links to it still.

Profilbild von Heather Cooper
Heather Coopervor 1 Jahr

Thank you :) I'm not aware of any place outside of requesting access. I have been a Google trusted tester for over a year now -NDA, etc So I'm lucky to get access quickly.

Ähnliche Videos

🎇 Video Model Comparison: Image2Video Video Model Comparison: Image2Video Same input image + text prompt on each model: • Pika 2.0 • PixVerse 3.5 • Runway Gen-3 • Kling AI 1.6 • Luma Dream Machine • Hailuo MiniMax I used the same Midjourney image and text prompt in each model and chose my favorite results.
1:00

🎇 Video Model Comparison: Image2Video Video Model Comparison: Image2Video Same input image + text prompt on each model: • Pika 2.0 • PixVerse 3.5 • Runway Gen-3 • Kling AI 1.6 • Luma Dream Machine • Hailuo MiniMax I used the same Midjourney image and text prompt in each model and chose my favorite results.

Heather Cooper

36,110 Aufrufe • vor 1 Jahr

🎬 Video Model Comparison: Text to video • Google Veo 2 • OpenAI Sora • Hunyuan Video • Pika 2.1 • Alibaba Wanx 2.1 • Runway Gen-3 • Kling 1.6 • Luma Ray2 • Hailuo T2V-01 I used the same prompt on each model for 2 generations and chose the best results below. Prompt: A majestic white tiger prowling silently through a dense jungle, sunlight piercing through the foliage to illuminate its fur; shot with a telephoto lens for depth, using soft natural lighting and creating a mysterious yet tranquil atmosphere. Prompt credit: LudovicCreator Overall, the results are impressive for each one. We have many choices for AI video! I could have tested more models, but this visual is crowded enough. Drop your results in the comments 👇
1:02

🎬 Video Model Comparison: Text to video • Google Veo 2 • OpenAI Sora • Hunyuan Video • Pika 2.1 • Alibaba Wanx 2.1 • Runway Gen-3 • Kling 1.6 • Luma Ray2 • Hailuo T2V-01 I used the same prompt on each model for 2 generations and chose the best results below. Prompt: A majestic white tiger prowling silently through a dense jungle, sunlight piercing through the foliage to illuminate its fur; shot with a telephoto lens for depth, using soft natural lighting and creating a mysterious yet tranquil atmosphere. Prompt credit: LudovicCreator Overall, the results are impressive for each one. We have many choices for AI video! I could have tested more models, but this visual is crowded enough. Drop your results in the comments 👇

Heather Cooper

255,315 Aufrufe • vor 1 Jahr

🔸Video Model Comparison: Image to video 6 Models included: • Pika 2.1 • Adobe Firefly • Runway Gen-3 • Kling 1.6 • Luma Ray2 • Hailuo T2V-01 This time I used an image generated with Magnific's new Fluid model ( Google DeepMind's Imagen + Mystic 2.5 ), and the same prompt in each model x 2 generations and chose my favorites below: "Low-angle dolly shot along the temple's edge, capturing the contrast between classical architecture and wild nature. Orange-tinted waves crash against rocks below as crimson clouds streak across the sky. Shot on anamorphic lenses, ultra-detailed textures, dramatic atmospheric lighting." I could have tested more models, but I didn't due to time constraints. Please feel free to reply with your own results!
0:43

🔸Video Model Comparison: Image to video 6 Models included: • Pika 2.1 • Adobe Firefly • Runway Gen-3 • Kling 1.6 • Luma Ray2 • Hailuo T2V-01 This time I used an image generated with Magnific's new Fluid model ( Google DeepMind's Imagen + Mystic 2.5 ), and the same prompt in each model x 2 generations and chose my favorites below: "Low-angle dolly shot along the temple's edge, capturing the contrast between classical architecture and wild nature. Orange-tinted waves crash against rocks below as crimson clouds streak across the sky. Shot on anamorphic lenses, ultra-detailed textures, dramatic atmospheric lighting." I could have tested more models, but I didn't due to time constraints. Please feel free to reply with your own results!

Heather Cooper

28,325 Aufrufe • vor 1 Jahr

🌊 Video Model Comparison: Image to Video → Google Veo 3 → Moonvalley Marey → Hailuo MiniMax 02 → Kling 2.1 → Midjourney V1 → Seedance Pro Image: Midjourney I ran the same prompt on each model 2-4 times and chose my favorite results below:
0:50

🌊 Video Model Comparison: Image to Video → Google Veo 3 → Moonvalley Marey → Hailuo MiniMax 02 → Kling 2.1 → Midjourney V1 → Seedance Pro Image: Midjourney I ran the same prompt on each model 2-4 times and chose my favorite results below:

Heather Cooper

67,548 Aufrufe • vor 1 Jahr

💄AI Video Model Comparison: Text to Video 10 Models included: • Google Veo2 • Adobe Firefly Video • Hunyuan Video • Pika 2.1 • Wan-2.1 • Runway Gen-3 • Kling AI 1.6 • Luma Ray2 • Hailuo MiniMax • Sora (not on grid, but in the full video) I used the same prompt on each model x 4 generations and chose my favorite results below. ⭕ Prompt: A person stands in a bright, airy room, bathed in soft natural light streaming through a large window adorned with sheer white curtains. As they gaze into a shimmering mirror, their hand, elegantly manicured with a deep crimson polish, lifts a glossy lip applicator towards their lips. Their outfit, a flowing pastel blouse paired with chic high-waisted trousers, exudes effortless style, complemented by delicate gold jewelry that sparkles subtly in the light. Applying lipstick, lip gloss, or any makeup seems to be extremely difficult for AI video models. It was hard to get a usable result for many of these. This prompt was generated on Nim with the Video Describe feature, and it worked better than my original prompt. Let me know what you think of the results!
1:10

💄AI Video Model Comparison: Text to Video 10 Models included: • Google Veo2 • Adobe Firefly Video • Hunyuan Video • Pika 2.1 • Wan-2.1 • Runway Gen-3 • Kling AI 1.6 • Luma Ray2 • Hailuo MiniMax • Sora (not on grid, but in the full video) I used the same prompt on each model x 4 generations and chose my favorite results below. ⭕ Prompt: A person stands in a bright, airy room, bathed in soft natural light streaming through a large window adorned with sheer white curtains. As they gaze into a shimmering mirror, their hand, elegantly manicured with a deep crimson polish, lifts a glossy lip applicator towards their lips. Their outfit, a flowing pastel blouse paired with chic high-waisted trousers, exudes effortless style, complemented by delicate gold jewelry that sparkles subtly in the light. Applying lipstick, lip gloss, or any makeup seems to be extremely difficult for AI video models. It was hard to get a usable result for many of these. This prompt was generated on Nim with the Video Describe feature, and it worked better than my original prompt. Let me know what you think of the results!

Heather Cooper

61,177 Aufrufe • vor 1 Jahr

✨AI Video Model Comparison: Image to Video • Pika 2.0 • Sora • Gen-3 • Kling 1.6 • Luma • Hailuo Image: Mystic 2.5 on Magnific.ai I used the same prompt on each model for 2-4 generations and picked my favorite results. Take a moment to watch the individual videos following the comparison grid - don't dismiss a result because of the lack of movement, without looking at the details, such as: Consistent color Quality of stone pathway Wind blowing grass & flowers Waves moving realistically below Lighthouse light flickering / moving Full workflow below:
1:01

✨AI Video Model Comparison: Image to Video • Pika 2.0 • Sora • Gen-3 • Kling 1.6 • Luma • Hailuo Image: Mystic 2.5 on Magnific.ai I used the same prompt on each model for 2-4 generations and picked my favorite results. Take a moment to watch the individual videos following the comparison grid - don't dismiss a result because of the lack of movement, without looking at the details, such as: Consistent color Quality of stone pathway Wind blowing grass & flowers Waves moving realistically below Lighthouse light flickering / moving Full workflow below:

Heather Cooper

69,089 Aufrufe • vor 1 Jahr

⚔️Video Model Comparison: Img2Vid I tested the same prompt on several models, but I was only able to get decent results from these: • Veo 2 • Ray2 • Kling 1.6 • Hailuo 01-Live Prompt: As the fierce warrior turns to rally his troops, sunlight gleams off his chiseled arms. In the foreground, he stands resolute, a harbinger of strength, while a dolly zoom reveals his warriors behind, poised with spears. Dust swirls around them under a vivid blue sky, capturing the impending surge of battle. Of course, I would tweak the prompt several times to get the motion I need, but I like to test multiple models to understand how they interpret prompts and images and the quality of the results. Image: Midjourney
0:47

⚔️Video Model Comparison: Img2Vid I tested the same prompt on several models, but I was only able to get decent results from these: • Veo 2 • Ray2 • Kling 1.6 • Hailuo 01-Live Prompt: As the fierce warrior turns to rally his troops, sunlight gleams off his chiseled arms. In the foreground, he stands resolute, a harbinger of strength, while a dolly zoom reveals his warriors behind, poised with spears. Dust swirls around them under a vivid blue sky, capturing the impending surge of battle. Of course, I would tweak the prompt several times to get the motion I need, but I like to test multiple models to understand how they interpret prompts and images and the quality of the results. Image: Midjourney

Heather Cooper

34,407 Aufrufe • vor 1 Jahr

🎇 Video Model Comparison: Text to Video → Veo 3 → Kling 2.1 Master → Hailuo 02 → Runway Gen-4 → Seedance 1 Pro → Pika 2.2 I compared the same prompt on each model x 2-4 generations and chose my favorites below: First-person perspective, 24mm lens simulation with motion blur and subtle head bob, bursting through a smashed window of a burning high-rise in hyperspeed as AI drones scan the area below; glass shards glint mid-air in bullet-time slow motion, sparks and rebar catching fire; wind howls past the camera with debris streaking across frame; gritty and kinetic, immersive audio and POV-style distortion, cinematic style. If you try it on a different video model, share your result below 👇
0:42

🎇 Video Model Comparison: Text to Video → Veo 3 → Kling 2.1 Master → Hailuo 02 → Runway Gen-4 → Seedance 1 Pro → Pika 2.2 I compared the same prompt on each model x 2-4 generations and chose my favorites below: First-person perspective, 24mm lens simulation with motion blur and subtle head bob, bursting through a smashed window of a burning high-rise in hyperspeed as AI drones scan the area below; glass shards glint mid-air in bullet-time slow motion, sparks and rebar catching fire; wind howls past the camera with debris streaking across frame; gritty and kinetic, immersive audio and POV-style distortion, cinematic style. If you try it on a different video model, share your result below 👇

Heather Cooper

64,523 Aufrufe • vor 1 Jahr

Video Model Comparison: Text to Video → Veo 3.1 → Sora 2 Pro → Kling 2.6 → Luma Ray3 → Wan 2.5 → Hailuo 2.3 → LTX-2 Pro → Seedance Pro Many of the new models automatically generate sound with video results including Veo 3.1, Sora 2 Pro, Kling 2.6, Wan 2.5, and LTX-2 Pro. I used Mirelo AI SFX and ElevenLabs to add sound to the others. I used the same prompt on each model 4x and chose the best result: Meteor Shower Panic Subject: [Father holding daughter's hand, looking up] Action: [Running through a suburban street, looking up as meteors streak across the sky, a small impact occurs nearby blowing wind and dust at them] Scene: [Suburban street, night, streetlights, burning trails in sky] Style: [Cinematic disaster, low angle tracking shot, high contrast, lens flares from meteors] Audio: Audio: sonic booms, wind rushing, car alarms triggered, shouting. Technical: [daytime, calm, sunny, happy] Which one is your favorite?
1:33

Video Model Comparison: Text to Video → Veo 3.1 → Sora 2 Pro → Kling 2.6 → Luma Ray3 → Wan 2.5 → Hailuo 2.3 → LTX-2 Pro → Seedance Pro Many of the new models automatically generate sound with video results including Veo 3.1, Sora 2 Pro, Kling 2.6, Wan 2.5, and LTX-2 Pro. I used Mirelo AI SFX and ElevenLabs to add sound to the others. I used the same prompt on each model 4x and chose the best result: Meteor Shower Panic Subject: [Father holding daughter's hand, looking up] Action: [Running through a suburban street, looking up as meteors streak across the sky, a small impact occurs nearby blowing wind and dust at them] Scene: [Suburban street, night, streetlights, burning trails in sky] Style: [Cinematic disaster, low angle tracking shot, high contrast, lens flares from meteors] Audio: Audio: sonic booms, wind rushing, car alarms triggered, shouting. Technical: [daytime, calm, sunny, happy] Which one is your favorite?

Heather Cooper

20,422 Aufrufe • vor 7 Monaten

Adobe just dropped a wild update at MAX... For the next 30 days, Creative Cloud + Firefly users get unlimited generations across Firefly and partner models like Gemini’s Nano Banana, Flux, Runway, and ChatGPT + Relaxed video generation I took the chance to test Firefly 5, their new model, with different types of photography, and the results literally blew my mind. Check out these Prompts and the link to the board to see for yourself. Don't forget to bookmark this This post is sponsored by Adobe as an Adobe Firefly Ambassador. #AdobeFireflyAmbassadors #Ad #FireflyPromo
0:56

Adobe just dropped a wild update at MAX... For the next 30 days, Creative Cloud + Firefly users get unlimited generations across Firefly and partner models like Gemini’s Nano Banana, Flux, Runway, and ChatGPT + Relaxed video generation I took the chance to test Firefly 5, their new model, with different types of photography, and the results literally blew my mind. Check out these Prompts and the link to the board to see for yourself. Don't forget to bookmark this This post is sponsored by Adobe as an Adobe Firefly Ambassador. #AdobeFireflyAmbassadors #Ad #FireflyPromo

Amira Zairi

44,767 Aufrufe • vor 8 Monaten

Mini tutorial for Google DeepMind 's VEO 2 Prompt Buddy. (Works well with other AI video tools as well). I’ve created my VEO 2 Prompt Buddy Custom GPT to help me go beyond my own brain and prompts. It can describe your images, enhance your prompts, be creative and more. It will provide you 4 different sections to choose from. I’ve also added a camera combinations section inspired by the Hailuo I2V-01-Director model for testing purposes. It’s currently in BETA mode in my GPT because, sometimes, multiple camera movements don’t work optimally in VEO 2 right now. Output sections: 1. PRO – Provides detailed, professional, and cinematically accurate results. 2. OUT OF THE BOX – Creates something more creative and unexpected from your idea. 3. ULTRA DETAILED – Expands your idea with extreme precision and even more detail. 4. CAMERA COMBINATIONS – Generates a version with cinematic camera transitions. This hot sauce is FREE for a limited time! Hope you have fun with this one. Link down below: --- All results are text-to-video
1:21

Mini tutorial for Google DeepMind 's VEO 2 Prompt Buddy. (Works well with other AI video tools as well). I’ve created my VEO 2 Prompt Buddy Custom GPT to help me go beyond my own brain and prompts. It can describe your images, enhance your prompts, be creative and more. It will provide you 4 different sections to choose from. I’ve also added a camera combinations section inspired by the Hailuo I2V-01-Director model for testing purposes. It’s currently in BETA mode in my GPT because, sometimes, multiple camera movements don’t work optimally in VEO 2 right now. Output sections: 1. PRO – Provides detailed, professional, and cinematically accurate results. 2. OUT OF THE BOX – Creates something more creative and unexpected from your idea. 3. ULTRA DETAILED – Expands your idea with extreme precision and even more detail. 4. CAMERA COMBINATIONS – Generates a version with cinematic camera transitions. This hot sauce is FREE for a limited time! Hope you have fun with this one. Link down below: --- All results are text-to-video

Allar Haltsonen

58,509 Aufrufe • vor 1 Jahr

Watch Battalion—My First Ever 5-Minute Gen AI Short Film. I had to upload it in two parts. This is part 1; part 2 is in the comments. I created this over a few weeks using 100% image-to-video and text-to-video Generative AI. While not perfect, the tools advance daily, even during production, helping to produce incredible results. This is my most in-depth and longest Gen AI narrative film project. It was created by one person. Watch the full-length 4K pro res here: Various tools helped me bring it to life, and it wouldn't have been possible without the power of AI. The story chronicles the 320th Barrage Balloon Batallion, the only unit to storm the beach during D-Day composed entirely of African American soldiers. They provided critical protection to the ships and soldiers below them from attacks by enemy aircraft. AI tools used: Kling AI Runway Luma Hailuo AI-MiniMax Hub Midjourney Magnific.ai OpenAI Stability AI udio FilmConvert Adobe Premiere Pro Adobe Photoshop Grok Comfyui, Live Portrait
2:22

Watch Battalion—My First Ever 5-Minute Gen AI Short Film. I had to upload it in two parts. This is part 1; part 2 is in the comments. I created this over a few weeks using 100% image-to-video and text-to-video Generative AI. While not perfect, the tools advance daily, even during production, helping to produce incredible results. This is my most in-depth and longest Gen AI narrative film project. It was created by one person. Watch the full-length 4K pro res here: Various tools helped me bring it to life, and it wouldn't have been possible without the power of AI. The story chronicles the 320th Barrage Balloon Batallion, the only unit to storm the beach during D-Day composed entirely of African American soldiers. They provided critical protection to the ships and soldiers below them from attacks by enemy aircraft. AI tools used: Kling AI Runway Luma Hailuo AI-MiniMax Hub Midjourney Magnific.ai OpenAI Stability AI udio FilmConvert Adobe Premiere Pro Adobe Photoshop Grok Comfyui, Live Portrait

Dave Clark

273,029 Aufrufe • vor 1 Jahr

So Runway Gen 4.5 finally adds image-to-video, the workflow most pros rely on for consistency. We put it head to head with Kling AI and Flow by Google VEO using the same reference images and prompts (below) to evaluate motion quality, stability, and cinematic realism. 1. Action/WaterPhysics Test Prompt: Cinematic, wide-shot of a man running in a shallow river. The camera is tracking the man from behind as he runs up the river. Handheld camera shake as the camera follows the man. 2. Fire Physics Test Prompt: Cinematic, wide-shot of terrified woman running towards her burning barn. She abruptly stops, and puts in hands on her head as she watches her barn burn down. 3. VFX test prompt: Cinematic, wide-shot of a hooded figure. Flashes of purple magic and smoke whirl around the figure. The figure lifts its arms as the purple magic and smoke intensifies. 4. 2D Animation Test Prompt: 2D animated shot of a waiting at a bus stop in a thunderstorm. The man turns, walks to the bench, and sits down. 5. 3D Animation Test Prompt: 3D animated shot of an octopus. The octopus reaches into a coral and picks up a glowing white gem. 6. Conversation Test Prompt: slow camera push-in as two friends are having a conversation at a coffee shop Overall verdict: Despite the “world’s best” claim, Runway Gen 4.5 is not there yet. Prompt adherence is solid, but motion, physics, and cinematic realism still lag behind tools like Kling and VEO. Great platform, mid-tier model for now.
0:47

So Runway Gen 4.5 finally adds image-to-video, the workflow most pros rely on for consistency. We put it head to head with Kling AI and Flow by Google VEO using the same reference images and prompts (below) to evaluate motion quality, stability, and cinematic realism. 1. Action/WaterPhysics Test Prompt: Cinematic, wide-shot of a man running in a shallow river. The camera is tracking the man from behind as he runs up the river. Handheld camera shake as the camera follows the man. 2. Fire Physics Test Prompt: Cinematic, wide-shot of terrified woman running towards her burning barn. She abruptly stops, and puts in hands on her head as she watches her barn burn down. 3. VFX test prompt: Cinematic, wide-shot of a hooded figure. Flashes of purple magic and smoke whirl around the figure. The figure lifts its arms as the purple magic and smoke intensifies. 4. 2D Animation Test Prompt: 2D animated shot of a waiting at a bus stop in a thunderstorm. The man turns, walks to the bench, and sits down. 5. 3D Animation Test Prompt: 3D animated shot of an octopus. The octopus reaches into a coral and picks up a glowing white gem. 6. Conversation Test Prompt: slow camera push-in as two friends are having a conversation at a coffee shop Overall verdict: Despite the “world’s best” claim, Runway Gen 4.5 is not there yet. Prompt adherence is solid, but motion, physics, and cinematic realism still lag behind tools like Kling and VEO. Great platform, mid-tier model for now.

Curious Refuge

25,538 Aufrufe • vor 5 Monaten

Watch California Monsters—My First Ever Google Veo 2 Full Film Trailer. This was mainly text-to-video. I handled the editing and sound design; Eleven Labs was used for voice, and Kling AI 1.6 was used for a few shots and lipsync parts. While not perfect, the tools advance every day, even during production, helping to produce insane results. One person created this over a weekend. Watch the full-length 4K pro res here: This was only possible with the power of Gen AI. The concept behind 'California Monsters' is a story I have always wanted to explore about what creatures might lurk in the fault lines that surround this beautiful state.
1:51

Watch California Monsters—My First Ever Google Veo 2 Full Film Trailer. This was mainly text-to-video. I handled the editing and sound design; Eleven Labs was used for voice, and Kling AI 1.6 was used for a few shots and lipsync parts. While not perfect, the tools advance every day, even during production, helping to produce insane results. One person created this over a weekend. Watch the full-length 4K pro res here: This was only possible with the power of Gen AI. The concept behind 'California Monsters' is a story I have always wanted to explore about what creatures might lurk in the fault lines that surround this beautiful state.

Dave Clark

89,963 Aufrufe • vor 1 Jahr

🧵1/3 I partnered with Kling to make a promo for their new 3.0 model. I came up with the concept, created it and delivered it all on my own in 3 days of early access, I wanted to make something that showed how Kling could be used to tell a diverse range of stories in a diverse range of styles. I've honestly been blown away be how incredible the model is and I haven't even scratched the surface of features like the new 3.0 OMNI model. For me Kling 2.6 was already best in class for most things but 3.0 sees massive improvements to video quality and detail (no need to add grain to mask imperfections) and prompt adherence. Also you can time your generations to be as long as 15 seconds and as short as 3, this is fantastic for pacing your scenes. Read on to find out more about the new Performance and Multi shot features!👇
1:26

🧵1/3 I partnered with Kling to make a promo for their new 3.0 model. I came up with the concept, created it and delivered it all on my own in 3 days of early access, I wanted to make something that showed how Kling could be used to tell a diverse range of stories in a diverse range of styles. I've honestly been blown away be how incredible the model is and I haven't even scratched the surface of features like the new 3.0 OMNI model. For me Kling 2.6 was already best in class for most things but 3.0 sees massive improvements to video quality and detail (no need to add grain to mask imperfections) and prompt adherence. Also you can time your generations to be as long as 15 seconds and as short as 3, this is fantastic for pacing your scenes. Read on to find out more about the new Performance and Multi shot features!👇

Uncanny Harry AI

28,977 Aufrufe • vor 5 Monaten

With the latest Seedance 2.0 release, there’s a feature we think might be even more transformative than the base video model itself: Seedance Omni. Similar to Kling Omni, Luma Modify, and Runway Aleph, Seedance Omni lets you guide the AI and make targeted edits to an existing video clip. It supports up to 9 reference images, 3 video clips, and 3 audio clips, allowing it to synthesize multiple layers of creative direction. We tested it across a range of scenarios (full prompts in the 🧵) 1. Modify eye color 2. Change weather 3. Time travel effect 4. Character swap 5. Add a spaceship 6. Change asteroids to meatballs 7. Dragon emerging from the clouds Verdict🏆: Seedance Omni excels at physical video dynamics, large visual effects, and environmental transformations. Its main weakness is resolution and output quality (around 720p), which can introduce flickering and softness.
1:02

With the latest Seedance 2.0 release, there’s a feature we think might be even more transformative than the base video model itself: Seedance Omni. Similar to Kling Omni, Luma Modify, and Runway Aleph, Seedance Omni lets you guide the AI and make targeted edits to an existing video clip. It supports up to 9 reference images, 3 video clips, and 3 audio clips, allowing it to synthesize multiple layers of creative direction. We tested it across a range of scenarios (full prompts in the 🧵) 1. Modify eye color 2. Change weather 3. Time travel effect 4. Character swap 5. Add a spaceship 6. Change asteroids to meatballs 7. Dragon emerging from the clouds Verdict🏆: Seedance Omni excels at physical video dynamics, large visual effects, and environmental transformations. Its main weakness is resolution and output quality (around 720p), which can introduce flickering and softness.

Curious Refuge

52,501 Aufrufe • vor 4 Monaten

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!
38:02

Apple built a large foundation model and fine-tuned it on multiple tasks. But they are doing something very clever: They load a single model in memory and use different adapters to specialize the model on the fly. I recorded a video to show you how to write the code to do the same thing Apple is doing. I explain everything step by step. Here is what I'll show you in the video: 1. We'll load two datasets 2. Then load a large model 3. Then, we'll fine-tune the model on both datasets I'll use LoRA to fine-tune the model. This process creates two small adapters, each specializing in solving one of the datasets. The base model's original parameters will remain unchanged. From here: 4. We'll generate a list of tasks 5. We'll load the correct adapter to solve each task The large model I'm using needs 346 MB of memory, but I only need to load it once. Each adapter is only 2.7 MB. I only need to load the base model once and pair it with any of the fine-tuned adapters. Minimum memory footprint and I can solve multiple tasks. Hope this helps!

Santiago

84,747 Aufrufe • vor 1 Jahr