Seedance|AI视频工坊
2.3K subscribers
3.96K photos
1.68K videos
1.9K links
Download Telegram
都进场了,看来国内 AI 硬件要爆发啊
This media is not supported in your browser
VIEW IN TELEGRAM
You can use all the same features with the new Product Placement tool that are available in HeyGen Avatar IV, like gesture control and uploading your own audio.

You can try it today on 's website!


🔞StableDiffusion一键生图,低操作门槛,画出你心目中的女神

📣 频道 @HotSora
I just uploaded an image of my character and my product and HeyGen generated a combined image featuring both, which you can then use to create the lip sync video.

Since the images are always in 9:16 format, I used Flux Kontext to convert it back to 16:9 for my video.
Hailuo AI's pricing is way cheaper than Veo 3, although it doesn’t support audio or lip sync yet.

It’s also great to see support for longer video generations (10s)

The 10s at 768p offers the best value at $0.52 per video, which gives you around 240 videos on the Ultra plan.


❤️角色上新-周思婷-你的前女友还在想着你-小黄文畅聊

📣 频道 @HotSora
This media is not supported in your browser
VIEW IN TELEGRAM
Not even Veo 3 could handle these prompts.

But Hailuo 2 shines when it comes to complex shots.

The coherence is impressive and the physics got a big upgrade.

Also, the first thing you’ll notice is how well Hailuo handles camera movements, their signature by now.
This media is not supported in your browser
VIEW IN TELEGRAM
If the leaderboard is true and Hailuo manages to ship it with native audio while keeping the unlimited plan, then I'm sure they will go to Number 1 real quick.

😍角色上新-你的女上司江婉如-纯爱轻M剧情

📣 频道 @HotSora
It seems like is actually about to dethrone Google Veo 3.

Knowing Minimax, they'll likely integrate an audio feature quickly (they of have to!).

Until it’s the standard, the first question people will ask is if it has native audio like Veo 3.
This media is not supported in your browser
VIEW IN TELEGRAM
It’s not just for products, you can also use the image editor to change literally everything else, like light, clothing, remove people or change whole environment.

You can try it today on 's website
Arcads AI launched a new feature called Prompt to AI UGC that allows you to generate or upload your own Avatars and create lip sync.

The best part is the new image editor!

You can upload your own objects to have your Actors/Avatars interact with your products:
This media is not supported in your browser
VIEW IN TELEGRAM
People are literally gaining MILLIONS of views with trending content like this.

So how can you ride the wave and use it to your advantage? The answer: Arcads AI

Here’s a quick test I made using my own avatar with lip sync holding my product:


😍NovelN女友AI-实习生搞定律所合伙人沈佳-职场纯爱

📣 频道 @HotSora
This media is not supported in your browser
VIEW IN TELEGRAM
Gesture control makes a huge difference, it makes the video feel more alive and gestures and movements are synced with the lip sync.

You can try it right now on 's website:
The new Avatar update from HeyGen supports 1080p resolution and introduces motion control, allowing you to guide gestures, facial expressions and even environmental movement.

You can just type how and what kind of motions you want in the prompt bar below:


👉一键推倒女高中生雨柔_NovelN女友机器人

📣 频道 @HotSora
Images generated using Flux Kontext Max. Videos made with 's new 2.1 model.

You can also just use all of them in
我去 Raycast 的 Windows 版本终于发布了

🤖AIBox新增特效-乳交特效-上传照片一键生成

> > >手冲实录< < <

📣 频道 @HotSora
This media is not supported in your browser
VIEW IN TELEGRAM
Stranger things V2 mini trailer 👻

--sref 3824510231
This media is not supported in your browser
VIEW IN TELEGRAM
新任牛马 Gemini Cli 第一个项目已经完工

让他写了个日签网站,搞得很快


> > >手冲实录< < <

📣 频道 @HotSora
刚刷到豆包的一个推送,他们自己用豆包 1.6 做了一下山东的2025 年高考题。

豆包Seed1.6-Thinking模型文科考了 683 分、立刻考了 648 分,理科跟 Gemini 并列第一。

今天山东公布了自己的高考录取分数线,基本可以稳定上 985 大学了,文科比 Gemini