Loading video...

Video Failed to Load

There was a problem loading this video. This could be due to a temporary network issue or the video might be unavailable.

Introducing Wan2.6 - A native multimodal model that turns your ideas into breathtaking videos and images! · Starring: Cast characters from reference videos into new scenes. Support human or human-like figures, enabling complex multi-person and human-object interactions with appearance and voice consistency. · Intelligent Multi-shot Narrative: Turn simple prompts... into auto-storyboarded, multi-shot videos. Maintain visual consistency and upgrade storytelling from single shots to rich narratives. · Native A/V Sync: Generate multi-speaker dialogue with natural lip-sync and studio-quality audio. It doesn’t just look real - it sounds real. · Cinematic Quality: 15s 1080p HD generation with comprehensive upgrades to instruction adherence, motion physics, and aesthetic control. · Advanced Image Synthesis and Editing: Deliver cinematic photorealism with precise control over lens and lighting. Support multi-image referencing for commercial-grade consistency and faithful aesthetic transfer. · Storytelling with Structure: Generate interleaved texts and images powered by real-world knowledge and reasoning capabilities, enabling hierarchical and structured visual narratives.show more

Wan

32,192 subscribers

3,848,218 views • 10 months ago •via X (Twitter)

Science & Technology

Anya Rossi• Live Now

Private livecam show

37 Comments

Wan10 months ago

The next era of multimodal AI for creators is here. Try it now and bring your stories to life!

Wan10 months ago

Join our Discord:

Greg Schoeninger10 months ago

Any plans on open sourcing?

L₳theesan10 months ago

lol, when they release Wan 2.5, they teased about open sourcing it "soon". With the new Wan 2.6, they didn't even mention it. It's giving SD3 🤣

sceptic10 months ago

Please open source

Tweeter10 months ago

No use to me if it's not open source... 😦

Sabir AI10 months ago

Impressive feature list but the real test is whether creators can get consistent results without fighting the model. Demos are easy, workflows are hard.

Wan10 months ago

try it with the free credit, and the model will speak for itself.

Sabir AI10 months ago

yes sure I definitely will try it

Shira Wiggles10 months ago

Is there a cpp plan?

Wenyang10 months ago

Trash. From the example on your website, I could feel the disappointment compared to wan2.5. Audio output is missing or terrible. Instruction-following is also awful. Do u put these models to show the advantages of your next generation?

Aaliya10 months ago

This is next-level multimodal storytelling, cinematic, consistent, and finally production-ready.

Ali Mamak10 months ago

Wan2.6 is aiming at the real bottleneck, temporal consistency and controllability across shots, not just pretty single clips. If it can lock identity, motion, and lip sync while taking reference casts into new scenes, that moves video generation from demos to production workflows. The next test is long form coherence, fast edit loops, and predictable control under constraints, that is where most models still break.

AI filmaker10 months ago

Tried it yesterday. Luckily it let me do it with free credits, unfortunately it was terrible . Wasn’t syncing . The audio came out double speed on 15 s clips. Can’t see the use or point of starting roll as You can’t use any other references , even then it’s added random mandarin speaking . It’s all over the shop!

Wan10 months ago

we'd want to know more details about this case, DM me directly plz

AI filmaker10 months ago

I’d love to but can’t work how !!

Altiam Kabir10 months ago

This sounds like a game changer for creators!

Saurabh Shukla10 months ago

It's a great day for video models. I hope the inference and training code is also available for these models

尚月10 months ago

源神，大家都在等你开源呀！

XRWorld10 months ago

Open source this or it did not happen!!! Very nasty from WAN team to play the open source song and then when people invested a lot of time and local resources into Wan 2.1/2.2... leave them with aged models that do not follow the prompts and are basically a WIP.

Aviv Ben Menachem10 months ago

awesome news, any chance for a new version soon on vace or animate?

Ⓜ️─█ Mamoute Media : Relais █─ 10 months ago

it V (BTS) in the women's version

Malzahran10 months ago

Wooow

AI/ML API10 months ago

Wan 2.6 is coming to AI/ML API today 👀

Pixazo10 months ago

Reference-based casting + native A/V sync is a massive step toward real narrative video, not just clips.

Michaelkaoi10 months ago

When will wan 2.6 on Qwen chat and 千問 app

ばいきんまん10 months ago

自分も地域で食糧を集めています！内容は良いですよ。

Tensor.Art10 months ago

Super excited about the upcoming event co hosted with official Wan team~😻 As exclusive strategic partner, Wan is the most popular models in the TensorArt. Our Wan community is full of amazing creators, cannot wait to try Wan2.6!🤩

Lydia_CanopyWave10 months ago

Love seeing multimodal models push toward real narrative control

Magica10 months ago

The fact that Wan 2.6 dropped and is already integrated into is lowkey the move... $15/month and I don't have to track releases anymore.

Mr Dari ₿10 months ago

Strong project here! Follow back 🔙 please 💎💯🔥

Min Chon Chi10 months ago

Complex multi-person interaction support is a welcome addition.

ODL10 months ago

Wan2.6 just raised the bar for creators 🔥, 1080p HD in 15s is insane 🤯, can't wait to see what stories come to life with this tech ✨

Bonosix10 months ago

用起来~

시원한술 haerin892 서울부터 부산까쥐!10 months ago

ww

Antaï Desena10 months ago

what prompt do you used

Rx Hasibur Rahman9 months ago

gm

Related Videos

Wan2.6: Commercial-Grade Image Output Wan2.6-Image is now available. 🖼️ Interleaved Text-and-Image Output：Generate interleaved text-and-image content with logical reasoning capabilities — enabling layered, narrative-driven visual storytelling. 🖼️ Multi-Image Conditioned Generation：Support flexible referencing, combining, and replacement of multiple images, integrating varied visual inspirations to generate novel and compelling results 🖼️ Commercial-Grade ID Preservation：High consistency in characters, styles, and elements for commercial scenarios. 🖼️ Extract creative elements — such as color, style, and composition — from reference images to enable aesthetically driven image generation. 🖼️ Precise Control of Camera Angles and Lighting：Support specifying camera perspective, spatial depth (foreground/background), and lighting details — enabling precise control over spatial composition and atmospheric mood. Try Wan2.6 today.

Wan2.6: Commercial-Grade Image Output Wan2.6-Image is now available. 🖼️ Interleaved Text-and-Image Output：Generate interleaved text-and-image content with logical reasoning capabilities — enabling layered, narrative-driven visual storytelling. 🖼️ Multi-Image Conditioned Generation：Support flexible referencing, combining, and replacement of multiple images, integrating varied visual inspirations to generate novel and compelling results 🖼️ Commercial-Grade ID Preservation：High consistency in characters, styles, and elements for commercial scenarios. 🖼️ Extract creative elements — such as color, style, and composition — from reference images to enable aesthetically driven image generation. 🖼️ Precise Control of Camera Angles and Lighting：Support specifying camera perspective, spatial depth (foreground/background), and lighting details — enabling precise control over spatial composition and atmospheric mood. Try Wan2.6 today.

Wan

33,310 views • 10 months ago

Wan2.6：Multi-shot Storytelling & Cinematic Shot Control ✨ Intelligent Storyboarding: Generate coherent videos with multiple shots from a single prompt. 🎬 Cross-Shot Consistency: Maintains a consistent main subject and scene across different shots. 🔊 Synchronized Audio: Generates video and synchronized sound simultaneously for a complete experience. ⏱️ Up to 15s Duration: Create longer videos, allowing for more dynamic shots and richer expression. Try Wan2.6 now at

Wan2.6：Multi-shot Storytelling & Cinematic Shot Control ✨ Intelligent Storyboarding: Generate coherent videos with multiple shots from a single prompt. 🎬 Cross-Shot Consistency: Maintains a consistent main subject and scene across different shots. 🔊 Synchronized Audio: Generates video and synchronized sound simultaneously for a complete experience. ⏱️ Up to 15s Duration: Create longer videos, allowing for more dynamic shots and richer expression. Try Wan2.6 now at

Wan

18,066 views • 10 months ago

Today, we launch Viggle V4 + Character Refine. - Strong character consistency with precise motion control - Upload 1 image → auto-generate multi-angle references - 1:1 motion transfer from reference videos - Multi-character swap in complex scenes - Generate in SECONDS at lower cost - Up to 60s videos Powered by our proprietary 3D foundation model, built for speed, reliability, and full creative control. Try it for free! 👇

Today, we launch Viggle V4 + Character Refine. - Strong character consistency with precise motion control - Upload 1 image → auto-generate multi-angle references - 1:1 motion transfer from reference videos - Multi-character swap in complex scenes - Generate in SECONDS at lower cost - Up to 60s videos Powered by our proprietary 3D foundation model, built for speed, reliability, and full creative control. Try it for free! 👇

ViggleAI

400,627 views • 7 months ago

Why do AI‑generated videos feel so real? Diffusion models turn noise into scenes that follow physics and human logic. With Wan AI model, creators control expressions, gestures, camera moves, and multi‑character action for cinematic realism. You imagine it. Wan brings it to life.

Why do AI‑generated videos feel so real? Diffusion models turn noise into scenes that follow physics and human logic. With Wan AI model, creators control expressions, gestures, camera moves, and multi‑character action for cinematic realism. You imagine it. Wan brings it to life.

Alibaba Cloud

3,476,792 views • 10 months ago

Today, we introduce HunyuanImage 3.0-Instruct, a native multimodal model focusing on image-editing by integrating visual understanding with precise image synthesis! 🚀 It understands input images and reasons before generating images. Built on an 80B-parameter MoE architecture (13B activated), it natively unifies deep multimodal comprehension and high-fidelity generation. 🧠 A "Thinking" Model with Native CoT & MixGRPO: The model doesn’t just execute commands, it processes them through a Native Chain-of-Thought (CoT) schema. Enhanced by our self-developed MixGRPO algorithm, it reasons through complex instructions to achieve flawless intent alignment and human-preference consistency. 🎨 Precise Editing & Multi-Image Fusion: The model enables accurate image editing by adding, removing, or modifying elements while keeping non-target areas perfectly intact. It also excels at seamless multi-image fusion, synthesizing complex scenes by extracting and blending elements from multiple sources into a unified, consistent output. 🏆 SOTA Performance: HunyuanImage 3.0-Instruct sets a new benchmark in visual quality and alignment, delivering performance that matches leading proprietary models. We aim to enable the community to explore new ideas with a state-of-the-art foundation model, fostering a dynamic and vibrant image generation ecosystem. 🛠️🎨 💻Try it at (PC only):

Today, we introduce HunyuanImage 3.0-Instruct, a native multimodal model focusing on image-editing by integrating visual understanding with precise image synthesis! 🚀 It understands input images and reasons before generating images. Built on an 80B-parameter MoE architecture (13B activated), it natively unifies deep multimodal comprehension and high-fidelity generation. 🧠 A "Thinking" Model with Native CoT & MixGRPO: The model doesn’t just execute commands, it processes them through a Native Chain-of-Thought (CoT) schema. Enhanced by our self-developed MixGRPO algorithm, it reasons through complex instructions to achieve flawless intent alignment and human-preference consistency. 🎨 Precise Editing & Multi-Image Fusion: The model enables accurate image editing by adding, removing, or modifying elements while keeping non-target areas perfectly intact. It also excels at seamless multi-image fusion, synthesizing complex scenes by extracting and blending elements from multiple sources into a unified, consistent output. 🏆 SOTA Performance: HunyuanImage 3.0-Instruct sets a new benchmark in visual quality and alignment, delivering performance that matches leading proprietary models. We aim to enable the community to explore new ideas with a state-of-the-art foundation model, fostering a dynamic and vibrant image generation ecosystem. 🛠️🎨 💻Try it at (PC only):

Tencent Hy

126,668 views • 9 months ago

Wan2.6: Be What You Wanna Be Wan2.6 R2V is now available. 🎬️Supports recording in real time or uploading a 5 second reference video, and replicating the person, animal, animated character, or object from that clip into new videos. 🎬️Supports both single- and multi-character video generation. 🎬️Replicates both the character’s appearance and voice, producing fully audio-visual synchronized output (including music, sound effects, and human speech).

Wan2.6: Be What You Wanna Be Wan2.6 R2V is now available. 🎬️Supports recording in real time or uploading a 5 second reference video, and replicating the person, animal, animated character, or object from that clip into new videos. 🎬️Supports both single- and multi-character video generation. 🎬️Replicates both the character’s appearance and voice, producing fully audio-visual synchronized output (including music, sound effects, and human speech).

Wan

28,910 views • 10 months ago

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

We've officially released and open-sourced HunyuanImage 2.1, our latest text-to-image model. The new model delivers on our commitment to balancing performance and quality. With native 2K image generation, HunyuanImage 2.1 is an advanced open-source text-to-image model.🎨 ✨ New in 2.1: 🔹Advanced Semantics: Supports ultra-long and complex prompts of up to 1000 tokens, and precisely controls the generation of multiple subjects in a single image. 🔹Precise Chinese and English Text Rendering with seamless image–text integration: The model naturally integrates text into images, making it suitable for a wide range of applications such as product covers, illustrations, and poster design to meet the needs of various fields. 🔹Rich Styles and High Aesthetic: Capable of generating images in various styles—including photorealistic portraits, comics, and vinyl figures—it delivers outstanding visual appeal and artistic quality. 🔹High-Quality Generation: Efficiently produces ultra-high-definition (2K) images in the same time other models take to generate a 1K image. HunyuanImage 2.1 uses two text encoders: a multimodal large language model (MLLM) to improve the model's image and text alignment capabilities, and a multi-language character-aware encoder to improve text rendering capabilities. The model is a single- and double-stream diffusion transformer with 17B parameters. We've also open-sourced the weights of the the accelerated version with meanflow which reduces inference steps from 100 to just 8, and PromptEnhancer, the first industrial-grade rewriting model that enhances your prompts for more nuanced and expressive image generation. Now, creators turn complex ideas—like posters with slogans or multi-panel comics—into visuals faster than ever. We’re just getting started. Stay tuned for our native multimodal image generation model coming soon. 🌐Website: 🔗Github: 🤗Hugging Face: ✨Hugging Face Demo:

Tencent Hy

89,430 views • 1 year ago

Big ideas deserve a powerful model. MiniMax H3 is now available on Fotor. From cinematic brand films and AI storytelling to ecommerce marketing, animation, and game-inspired visuals, MiniMax H3 is built to handle every stage of your creative workflow. With native multimodal understanding, multi-reference generation, advanced character and camera control, and expressive voice capabilities, turning complex ideas into production-ready content has never been easier. One model. Endless creative possibilities. Try MiniMax H3 on Fotor today.

Fotor

81,660 views • 2 months ago

MagicScroll: Nontypical Aspect-Ratio Image Generation for Visual Storytelling via Multi-Layered Semantic-Aware Denoising paper page: Visual storytelling often uses nontypical aspect-ratio images like scroll paintings, comic strips, and panoramas to create an expressive and compelling narrative. While generative AI has achieved great success and shown the potential to reshape the creative industry, it remains a challenge to generate coherent and engaging content with arbitrary size and controllable style, concept, and layout, all of which are essential for visual storytelling. To overcome the shortcomings of previous methods including repetitive content, style inconsistency, and lack of controllability, we propose MagicScroll, a multi-layered, progressive diffusion-based image generation framework with a novel semantic-aware denoising process. The model enables fine-grained control over the generated image on object, scene, and background levels with text, image, and layout conditions. We also establish the first benchmark for nontypical aspect-ratio image generation for visual storytelling including mediums like paintings, comics, and cinematic panoramas, with customized metrics for systematic evaluation. Through comparative and ablation studies, MagicScroll showcases promising results in aligning with the narrative text, improving visual coherence, and engaging the audience. We plan to release the code and benchmark in the hope of a better collaboration between AI researchers and creative practitioners involving visual storytelling.

MagicScroll: Nontypical Aspect-Ratio Image Generation for Visual Storytelling via Multi-Layered Semantic-Aware Denoising paper page: Visual storytelling often uses nontypical aspect-ratio images like scroll paintings, comic strips, and panoramas to create an expressive and compelling narrative. While generative AI has achieved great success and shown the potential to reshape the creative industry, it remains a challenge to generate coherent and engaging content with arbitrary size and controllable style, concept, and layout, all of which are essential for visual storytelling. To overcome the shortcomings of previous methods including repetitive content, style inconsistency, and lack of controllability, we propose MagicScroll, a multi-layered, progressive diffusion-based image generation framework with a novel semantic-aware denoising process. The model enables fine-grained control over the generated image on object, scene, and background levels with text, image, and layout conditions. We also establish the first benchmark for nontypical aspect-ratio image generation for visual storytelling including mediums like paintings, comics, and cinematic panoramas, with customized metrics for systematic evaluation. Through comparative and ablation studies, MagicScroll showcases promising results in aligning with the narrative text, improving visual coherence, and engaging the audience. We plan to release the code and benchmark in the hope of a better collaboration between AI researchers and creative practitioners involving visual storytelling.

AK

22,379 views • 2 years ago

We’re bringing identity consistency to Veo 3.1. For the first time ever, you have complete control of your Veo generations across every voice, character, and scene. With HeyGen’s avatar technology, generate once in Veo 3.1 and your avatar stays consistent, keeping your real voice and true look every time. Voice Consistency: Choose your voice, and HeyGen blends your real tone into Veo’s cinematic motion. Multi-Scene Consistency: Seamless transitions between scenes, one identity, endless stories. Character Consistency: Keep using that avatar with consistent voice and appearance to create unlimited videos in different outfits, styles and environments This is what happens when the world’s best video generator meets the best avatar technology. Don’t just make a scene. Free your story.

HeyGen

1,864,420 views • 1 year ago

1/2 Meet Wan2.7-Video — The Comprehensive Model for Controllable Video Storytelling! From single clips to full-scale narrative direction, we’ve built more than just a generator. We’ve built a director’s suite: • Multimodal control over performance and style via text, image, audio, and video. • Character customization with up to 5 reference inputs and voice profiles. • Video editing with simple, intuitive instructions. • Full-stack creative toolkit: generation, editing, cloning, restyling, continuation, and more. • Sustained improvements in visual fidelity, motion stability, and prompt adherence.

1/2 Meet Wan2.7-Video — The Comprehensive Model for Controllable Video Storytelling! From single clips to full-scale narrative direction, we’ve built more than just a generator. We’ve built a director’s suite: • Multimodal control over performance and style via text, image, audio, and video. • Character customization with up to 5 reference inputs and voice profiles. • Video editing with simple, intuitive instructions. • Full-stack creative toolkit: generation, editing, cloning, restyling, continuation, and more. • Sustained improvements in visual fidelity, motion stability, and prompt adherence.

Wan

25,573,373 views • 6 months ago

HappyHorse 1.1 is now available in ComfyUI as a Partner Node. Native audio. Multi-character consistency. Cinematic framing. All in one render pass. What's new in 1.1: → Dynamic expressiveness: Smoother motion, no more stiff movement → Enhanced R2V: Up to 9 reference images per generation → Multi-character consistency: Each character holds their look, zero cross-contamination → Flexible char × scene: Characters stay consistent as backgrounds change → Better instruction following: Handles 2,500+ char prompts, 6–8 scenes in one pass → Natural skin: Fixes shiny skin and over-sharpening for close-up commercial work → Cinematic language: Shot-reverse-shot, tracking shots, cohesive transitions → Upgraded audio: Tighter dialogue + SFX sync with emotional performance layered in Click the link below for the blog & the workflow 👇

HappyHorse 1.1 is now available in ComfyUI as a Partner Node. Native audio. Multi-character consistency. Cinematic framing. All in one render pass. What's new in 1.1: → Dynamic expressiveness: Smoother motion, no more stiff movement → Enhanced R2V: Up to 9 reference images per generation → Multi-character consistency: Each character holds their look, zero cross-contamination → Flexible char × scene: Characters stay consistent as backgrounds change → Better instruction following: Handles 2,500+ char prompts, 6–8 scenes in one pass → Natural skin: Fixes shiny skin and over-sharpening for close-up commercial work → Cinematic language: Shot-reverse-shot, tracking shots, cohesive transitions → Upgraded audio: Tighter dialogue + SFX sync with emotional performance layered in Click the link below for the blog & the workflow 👇

ComfyUI

17,266 views • 4 months ago

Kling 4.0 is coming this October. Kling 4.0 Flash is live now for Ultra Yearly subscribers. Kling 4.0 brings video creation to a new level of visual realism, creative control, and narrative completeness. 🎞️ Upgraded Audio & Visuals: Stable dynamic motion, high-quality stereo audio, more accurate lip sync, up to 4K resolution, and 10-bit HDR output. 🔮 Omni Reference: Richer reference options with up to 15 multi-modal references, more consistent results, and enhanced video editing. 🎥 Seamless Storytelling: Multi-keyframe control supporting up to 10 keyframes, native 30-second generation. 🌍 Diverse Possibilities: Video extension, support for multiple languages, accents, and dialects. The stage is set. You call the shots!

Kling 4.0 is coming this October. Kling 4.0 Flash is live now for Ultra Yearly subscribers. Kling 4.0 brings video creation to a new level of visual realism, creative control, and narrative completeness. 🎞️ Upgraded Audio & Visuals: Stable dynamic motion, high-quality stereo audio, more accurate lip sync, up to 4K resolution, and 10-bit HDR output. 🔮 Omni Reference: Richer reference options with up to 15 multi-modal references, more consistent results, and enhanced video editing. 🎥 Seamless Storytelling: Multi-keyframe control supporting up to 10 keyframes, native 30-second generation. 🌍 Diverse Possibilities: Video extension, support for multiple languages, accents, and dialects. The stage is set. You call the shots!

Kling AI

2,477,507 views • 25 days ago

Dreamina Seedance 2.0 is LIVE on CapCut app, desktop & web, starting gradually in Indonesia, Philippines, Thailand, Vietnam, Malaysia, Brazil and Mexico with expansion over time. Generate and edit with industry-leading quality in one seamless workflow. Built to unlock new possibilities in visual storytelling with CapCut: - SOTA long-form coherence, up to 15s videos with multi-shot storytelling and exceptional text prompt adherence - Built-in dialogue, lipsync, and immersive spatial sound - Multimodal reference for greater creative control and precision Find Dreamina Seedance 2.0 in the following CapCut features: - Quick try: AI Lab & AI Generator (app, v17.1.0) - Generate + edit workflow: Media → AI Video (app & desktop) - AI-native workflow with omni reference: Video Studio (our latest canvas-based ai production workspace built for everyone from beginners to pro, access via CapCut web)

Dreamina Seedance 2.0 is LIVE on CapCut app, desktop & web, starting gradually in Indonesia, Philippines, Thailand, Vietnam, Malaysia, Brazil and Mexico with expansion over time. Generate and edit with industry-leading quality in one seamless workflow. Built to unlock new possibilities in visual storytelling with CapCut: - SOTA long-form coherence, up to 15s videos with multi-shot storytelling and exceptional text prompt adherence - Built-in dialogue, lipsync, and immersive spatial sound - Multimodal reference for greater creative control and precision Find Dreamina Seedance 2.0 in the following CapCut features: - Quick try: AI Lab & AI Generator (app, v17.1.0) - Generate + edit workflow: Media → AI Video (app & desktop) - AI-native workflow with omni reference: Video Studio (our latest canvas-based ai production workspace built for everyone from beginners to pro, access via CapCut web)

CapCut

7,174,407 views • 7 months ago

Been playing with a character consistency workflow again. This came from Nexora that starts with a single GPT Image 2.0 character sheet and turns it into a multi-shot cinematic sequence in Seedance 2.0. One thing I added to the workflow: I first comped the character, environment, and visual style together in Midjourney before bringing everything into GPT Image 2.0 to build the character sheet. I also pushed it a bit further by generating multiple characters, destinations, and art directions at the same time, then turning those into production-style character sheets and shot sequences. It's interesting how much consistency you can get from a single well-defined character board when the visual language, wardrobe, and environment are established upfront. Check out three different character + scenes here 👇

Been playing with a character consistency workflow again. This came from Nexora that starts with a single GPT Image 2.0 character sheet and turns it into a multi-shot cinematic sequence in Seedance 2.0. One thing I added to the workflow: I first comped the character, environment, and visual style together in Midjourney before bringing everything into GPT Image 2.0 to build the character sheet. I also pushed it a bit further by generating multiple characters, destinations, and art directions at the same time, then turning those into production-style character sheets and shot sequences. It's interesting how much consistency you can get from a single well-defined character board when the visual language, wardrobe, and environment are established upfront. Check out three different character + scenes here 👇

Curious Refuge

12,209 views • 4 months ago

🚀 Introducing the Kling 3.0 Model: Everyone a Director. It’s Time. An all-in-one creative engine that enables truly native multimodal creation. - Superb Consistency: Your characters and elements, always locked in. - Flexible Video Production: Create 15s clips with precise control, excellent video realism and customizable multi-shots. - Upgraded Native Audio: Now supports multi-character reference, more languages and accents. - Enhanced Image Generation: 4K Image output, new image series mode and more cinematic visuals. Ultra subscribers get exclusive early access — now live on Web -> Kling AI. Get ready to unlock your creative potential with our most advanced model yet. [For the next 24 hours ONLY] Follow, Comment & Retweet to win exclusive early access - we will select 10 winners in this post from the top ten 'Most liked replies'. So which feature are you most excited about? Let us know in the comment! We will contact you via your DM.

🚀 Introducing the Kling 3.0 Model: Everyone a Director. It’s Time. An all-in-one creative engine that enables truly native multimodal creation. - Superb Consistency: Your characters and elements, always locked in. - Flexible Video Production: Create 15s clips with precise control, excellent video realism and customizable multi-shots. - Upgraded Native Audio: Now supports multi-character reference, more languages and accents. - Enhanced Image Generation: 4K Image output, new image series mode and more cinematic visuals. Ultra subscribers get exclusive early access — now live on Web -> Kling AI. Get ready to unlock your creative potential with our most advanced model yet. [For the next 24 hours ONLY] Follow, Comment & Retweet to win exclusive early access - we will select 10 winners in this post from the top ten 'Most liked replies'. So which feature are you most excited about? Let us know in the comment! We will contact you via your DM.

Kling AI

1,197,986 views • 8 months ago

DisCo: Disentangled Control for Referring Human Dance Generation in Real World paper page: Generative AI has made significant strides in computer vision, particularly in image/video synthesis conditioned on text descriptions. Despite the advancements, it remains challenging especially in the generation of human-centric content such as dance synthesis. Existing dance synthesis methods struggle with the gap between synthesized content and real-world dance scenarios. In this paper, we define a new problem setting: Referring Human Dance Generation, which focuses on real-world dance scenarios with three important properties: (i) Faithfulness: the synthesis should retain the appearance of both human subject foreground and background from the reference image, and precisely follow the target pose; (ii) Generalizability: the model should generalize to unseen human subjects, backgrounds, and poses; (iii) Compositionality: it should allow for composition of seen/unseen subjects, backgrounds, and poses from different sources. To address these challenges, we introduce a novel approach, DISCO, which includes a novel model architecture with disentangled control to improve the faithfulness and compositionality of dance synthesis, and an effective human attribute pre-training for better generalizability to unseen humans. Extensive qualitative and quantitative results demonstrate that DISCO can generate high-quality human dance images and videos with diverse appearances and flexible motions.

DisCo: Disentangled Control for Referring Human Dance Generation in Real World paper page: Generative AI has made significant strides in computer vision, particularly in image/video synthesis conditioned on text descriptions. Despite the advancements, it remains challenging especially in the generation of human-centric content such as dance synthesis. Existing dance synthesis methods struggle with the gap between synthesized content and real-world dance scenarios. In this paper, we define a new problem setting: Referring Human Dance Generation, which focuses on real-world dance scenarios with three important properties: (i) Faithfulness: the synthesis should retain the appearance of both human subject foreground and background from the reference image, and precisely follow the target pose; (ii) Generalizability: the model should generalize to unseen human subjects, backgrounds, and poses; (iii) Compositionality: it should allow for composition of seen/unseen subjects, backgrounds, and poses from different sources. To address these challenges, we introduce a novel approach, DISCO, which includes a novel model architecture with disentangled control to improve the faithfulness and compositionality of dance synthesis, and an effective human attribute pre-training for better generalizability to unseen humans. Extensive qualitative and quantitative results demonstrate that DISCO can generate high-quality human dance images and videos with diverse appearances and flexible motions.

AK

161,479 views • 3 years ago

Learn about Google’s new SOTA image model, Gemini 2.5 Flash, its key capabilities, and what’s next on the roadmap with some of the team behind the model Nicole Brichtova Kaushik Shivakumar Mostafa Dehghani Robert Riachi with Logan Kilpatrick. Timecodes: 0:37 New model introduction 01:21 Demo: Image editing 03:44 Text rendering capabilities 04:44 Beyond human preference evals 06:44 Text rendering as a proxy for quality 08:38 Positive transfer between modalities 11:25 Demo: multi-turn, context aware image generation 13:54 Pixel-perfect editing and character consistency 15:51 Interleaved image generation 17:59 Specialized vs. native models 19:52 Understanding nuanced prompts 20:59 User feedback shaping model development 22:37 Improvements in character consistency 24:17 More natural looking images from team collaboration 26:41 What’s next for image generation models

Learn about Google’s new SOTA image model, Gemini 2.5 Flash, its key capabilities, and what’s next on the roadmap with some of the team behind the model Nicole Brichtova Kaushik Shivakumar Mostafa Dehghani Robert Riachi with Logan Kilpatrick. Timecodes: 0:37 New model introduction 01:21 Demo: Image editing 03:44 Text rendering capabilities 04:44 Beyond human preference evals 06:44 Text rendering as a proxy for quality 08:38 Positive transfer between modalities 11:25 Demo: multi-turn, context aware image generation 13:54 Pixel-perfect editing and character consistency 15:51 Interleaved image generation 17:59 Specialized vs. native models 19:52 Understanding nuanced prompts 20:59 User feedback shaping model development 22:37 Improvements in character consistency 24:17 More natural looking images from team collaboration 26:41 What’s next for image generation models

Google AI Developers

31,745 views • 1 year ago

Listen up 🔊 We’ve made some updates to our Gemini Audio models and capabilities: — Gemini’s live speech-to-speech translation capability is rolling out in a beta experience to the Google Translate app, bringing you real-time audio translation that captures the nuance of human speech — Gemini 2.5 Flash and 2.5 Pro Text-to-Speech preview models bring improved adherence to style prompts, precision pacing with context-aware speed adjustments, and character voice consistency for multi-speaker scenarios — Gemini 2.5 Flash Native Audio is now updated, with improvements to handle complex workflows, navigate user instructions, and hold natural conversations

Listen up 🔊 We’ve made some updates to our Gemini Audio models and capabilities: — Gemini’s live speech-to-speech translation capability is rolling out in a beta experience to the Google Translate app, bringing you real-time audio translation that captures the nuance of human speech — Gemini 2.5 Flash and 2.5 Pro Text-to-Speech preview models bring improved adherence to style prompts, precision pacing with context-aware speed adjustments, and character voice consistency for multi-speaker scenarios — Gemini 2.5 Flash Native Audio is now updated, with improvements to handle complex workflows, navigate user instructions, and hold natural conversations

Google AI

1,095,303 views • 10 months ago