Explore Seedance Creations

I've used a new ending for Breaking Bad using Seedance 2.0

I must say this model is truly amazing. It's the most impressive I've ever experienced.

Contextual AI Raises $80 Million, Judge Calls Google A Monopolist, Bytedance Intros Jimeng Text-To-Video AI

ly shut down, but X also named CVS, Unilever, Mars and the Danish energy company Ørsted as defendants in the suit. His suit against the non-profit progressive watchdog group Media Matters, which publicized what it found to be hate-speech on X, goes to court in March.

ByteDance Enters the AI Text-to-Video Generation Space with Jimeng AI 📽️

ByteDance, a Chinese tech giant behind TikTok and the viral video editor CapCut, has launched its first AI text-to-video model, Jimeng AI. Currently exclusive to China, this model shows ByteDance’s interest in entering the increasingly competitive field of AI video generation. Jimeng AI could be a potential rival to OpenAI’s upcoming Sora model. # Key Features of Jimeng AI: 1. **Jimeng AI’s Development**: * The model was developed by Faceu Technology, a ByteDance subsidiary known for the CapCut video editing app. 1. **Comparative Landscape**: * While ByteDance is a significant player, it’s not alone in this space. Kuaishou, another major Chinese video platform, recently made its AI text-to-video tool, Kling AI, available outside China. 1. **Capabilities of Jimeng AI**: * Jimeng AI can generate brief AI video content from text or images like modern AI video tools. * It also offers control over camera movements and the framing of the video’s first and last scenes. 1. **Focus on Chinese Text and Culture**: * Faceu has prioritized making Jimeng AI adept at understanding and following Chinese text prompts. 1. **Market Positioning**: * Jimeng AI is mobile-first and tied closely to social video platforms like Douyin and CapCut. Read more: [https://aitoolsclub.com/bytedance-enters-the-ai-text-to-video-generation-space-with-jimeng-ai/](https://aitoolsclub.com/bytedance-enters-the-ai-text-to-video-generation-space-with-jimeng-ai/) https://preview.redd.it/b2lalo7nneid1.png?

Watching AI Films On A Cinema Screen Exposed The Tech’s Huge Flaws — But Seedance 2.0 Showed Things Can Change Fast

Hulk vs Thor — and this entire fight was generated by Seedance 2.0

You can’t skip this. What you’re watching is **Hulk vs Thor** — recreated entirely with **Seedance 2.0**. This isn’t a movie clip. This isn’t studio VFX. This is AI-generated video with real cinematic motion. The impact of Hulk’s punch. Thor blocking with Mjolnir. The camera movement, timing, and weight behind every hit. It genuinely feels like a finished Marvel-level action scene, not an AI experiment. If Seedance 2.0 can already pull off character-to-character combat like this, the future of AI filmmaking is going to get wild very fast. Curious what you think?

Jia Zhangke’s Seedance 2.0 Short Film

I’ve promised multiple times on this sub that the second I saw an AI short that I actually thought had any merit, I’d post it. Here it is. It’s odd because I’m someone most people on here would call an “anti”, I’m generally skeptical and I think even people who are in love with AI should be vigilant of its limitations and exploitations. I wouldn’t change most of what I have to say after seeing this short. One of the greatest living artists tried it and the fundamental limitations are still there. Here’s what makes me feel even weirder, though: this short hasn’t been posted here yet. It was released days ago. Jia is one of the greatest filmmakers working, and possibly the greatest mainland Chinese filmmaker ever. There are thousands of users here who will send me the most garish, uncanny “short films” in a bid to prove AI filmmaking’s legitimacy. Then an actually good AI short was released by a major filmmaker and they don’t even seem to know it exists. So that’s my victory lap I guess, this is definitive proof that I won’t just say something is bad because it’s AI. Somehow I actually did a better job seeking out good work than any of the “pros”. Perhaps they should reflect on why that is.

Seedance fighting the future of truth?

I'm sure everyone has seen the Tom Cruise and Brad Pitt fight that was created with Seedance 2.0. It's impressive for many reasons. But very curious what this sub thinks about what it means for the future of truth verification. Growing up it was pics or it didn't happen. Then it was video. Now both are (or will soon be able to be) faked easily. Thoughts? [https://www.youtube.com/watch?v=23qpkGQdi1g](https://www.youtube.com/watch?v=23qpkGQdi1g)

Seedance 2.0 just directed a full emotional film scene from text

**Seedance 2.0 is insane.** This was generated from a single prompt. Full cinematic scene. Multiple cuts. Emotional performance. Natural camera direction. **Prompt used for this video:** “A multi-shot cinematic sequence. Starts with a wide shot. The man leans against a tree, touching his hair and looking upward. The woman watches him quietly. Cut to a close-up as he looks at the canopy, then turns away. Cut back wide as she walks toward him. The camera tracks her. She places her hand on his shoulder. He makes eye contact and breaks down crying. Heavy breathing. Rustling leaves. Moss footsteps.” AI is learning filmmaking language. This changes everything. Let me know your thoughts in the comment section below!

Seedance 2.0 turned cats into comedy gold

I honestly can’t get over how good (and funny) these Seedance 2.0 cat videos are. The expressions, the timing, the little movements — it feels way too real. This one had me laughing instantly. AI videos used to feel stiff and now they’re just pure chaos in the best way.

Where’s Waldo?" Heist Trailer — 60 hours of work using Seedance 1.5 Pro, Higgsfield 4K Skin Enhance, and 11 Labs

Is the era of the "rubbery" AI look finally over? I pushed **Seedance 1.5 Pro** and **Higsfield AI** to the absolute limit to see if I could create a hyperrealistic heist trailer for the Waldo franchise. From 4K skin textures to custom voice cloning in **Eleven Labs**, I’m taking you "under the hood" to show you the exact workflow I used to bridge the uncanny valley. If you’re tired of generic AI video generations and want to learn how to get your "actors" to actually act, this one is for you. ⚠️ **DISCLAIMER:** This project was created purely for educational purposes to test the limits of current AI video and voice models. All celebrity likenesses and voices were generated using AI tools for demonstration and research into hyperrealistic workflows. 🎬 **WATCH THE FULL TUTORIAL:** I’ve posted the complete 60-hour breakdown and step-by-step workflow on my YouTube channel: [https://www.youtube.com/watch?v=Z5OWXIzMZdM](https://www.youtube.com/watch?v=Z5OWXIzMZdM)

Seedance 2.0 gives GTA: San Andreas an AI glow-up — and it feels like the future

This honestly feels unreal to watch. What you’re seeing here is **GTA: San Andreas**, reimagined using **Seedance 2.0** and modern AI workflows. The environments, lighting, and overall atmosphere feel closer to a next-gen remake than a game that originally came out in 2004. What really stands out isn’t just sharper visuals — it’s the *vibe*. The streets still feel like San Andreas. The scale, the mood, the nostalgia… it’s all there. But now it’s wrapped in cinematic lighting, realistic shadows, and a level of depth that makes it feel alive again. This doesn’t feel like a simple filter or a mod. It feels like AI is being used as a **creative reconstruction tool**, rebuilding a classic from scratch while respecting what made it iconic in the first place. Seeing something this familiar transformed so convincingly really makes you think about the future of gaming: * AI remasters instead of traditional remakes * Classic worlds preserved, upgraded, and re-experienced * Entire generations rediscovering old games in a new form Honestly, this era of gaming is something else. Would you play a full GTA: San Andreas experience rebuilt like this using AI?

[Comparison] Seedance 2.0 vs Kling 3.0 & Veo 3 — Is 2.0 Already Ahead?