受邀参与了一个海外设计 Agents 产品 lovart ai 的测试
试了一下这也太 tm 牛皮了！
我只说了 50 个字的需求，直接给我生成了模特 ID 图片、11 个视频分镜图片，每个分镜的拍摄指导和分镜视频。
最后直接自动剪辑成了完整的视频！
下面是一些详细的介绍：  https://twitter.com/lovart_ai/status/1921958554312831133

Twitter

I was invited to test a <lovart ai> overseas design Agents product
I tried it and it's so damn awesome!
I only gave a 50-word requirement, and it directly generated model ID images, 11 video storyboard images, shooting guidance for each scene, and scene videos.
Finally, it automatically edited a complete video!
Below are some detailed introductions: https://twitter.com/lovart_ai/status/1921958554312831133

Crypto traders have not yet shown enough fear on social media to confirm a market bottom, according to a crypto analyst who suggested Bitcoin could still slide to around $75,000.
“It looks very tempti...

Crypto market ‘isn’t scared enough’ to call a bottom yet: Santiment

Author | Shisi Jun

1. Introduction
In the blink of an eye, I've been working in the wallet industry for four years. Many people think the wallet landscape in 2025 is already solidified, but that's not the case—it's been turbulent beneath the surface. This year:

· Coinbase launched its new CDP wallet, built on TEE technology;

· Binance's MPC wallet introduces key sharding and hosting to the TEE environment;

· Bitget just released its social login feature last week, hosted on a TEE basis;

· ...

A brief overview of the history of blockchain wallets and the market landscape in 2025.

According to ME News, on December 21st (UTC+8), Ethereum founder Vitalik Buterin posted on Farcaster, stating that prediction markets are a good remedy for volatile opinions on emotional topics. He used two screenshots as examples: Elon Musk previously posted that a British civil war was "inevitable," while the prediction market Polymarket showed only a 3% probability of a "Will there be a British civil war in 2024?" (Vitalik believes 3% is still too high, as some bettors have inflated the probability). Vitalik...

Vitalik: Prediction markets are the "cure" for social media, deflecting the wild opinions of emotional topics.

主要的功能有：
在一个产品里面使用GPT image-1、 Flux pro 、OpenAI-o3、Gemini Imagen 3、Kling AI、Tripo AI 、Suno AI 等你能想到的任何图像、视频、音频模型
只需要提出一次提示词他会自动帮你完成从策划到设计到执行的所有过程，批量出图和视频
除了正常跟 Agents 

The main functions are:
Use GPT image-1, Flux pro, OpenAI-o3, Gemini Imagen 3, Kling AI, Tripo AI, Suno AI and any other image, video, and audio models you can think of in one product
You only need to give a prompt word once and it will automatically help you complete the entire process from planning to design to execution, and batch output of images and videos
In addition to normal communication with Agents

详细看一下我这个例子：
我发的提示词是“帮我为这个产品生成一个 30 秒广告需要的所有分镜图片，后续我会基于这些图片指导拍摄和生成视频”
然后他先自己生成了一个模特的全身图用来参考
然后自己编排提示词生成了 11 张分镜图， 每个图片里面的人物都是一致的 

Take a closer look at my example:
The prompt I sent was "Help me generate all the storyboards for a 30-second ad for this product. I will guide the shooting and video generation based on these pictures later."
Then he first generated a full-body picture of a model for reference
Then he arranged the prompts and generated 11 storyboards, and the characters in each picture were the same

这里可以看到这个任务的回放：https://www.lovart.ai/r/8nn2vxz

A replay of this mission can be seen here: https://www.lovart.ai/r/8nn2vxz

另外除了视频之外 Lovart AI 对图片素材生产提效也很大
比如这里我让他把右边的白色 T 恤穿在图 1 的模特身上，然后基于这个图片一次性生成了不同平台需要的广告图，我全程没有干预，太强了。 

In addition to videos, Lovart AI also greatly improves the efficiency of image material production. For example, here I asked it to put the white T-shirt on the model in Figure 1, and then based on this picture, it generated the advertising pictures required by different platforms at one time. I did not intervene in the whole process. It is so powerful.

A one-picture introduction of the product