This media is not supported in your browser
VIEW IN TELEGRAM
Animation generation in After Effects using GPT-5.4.
The Atom plugin is used here. https://tryatom.ai/
Under the hood, it runs MCP connected to After Effects.
After Effects also has a version with Gemini and lots of other fun stuff.
The Atom plugin is used here. https://tryatom.ai/
Under the hood, it runs MCP connected to After Effects.
After Effects also has a version with Gemini and lots of other fun stuff.
๐365๐107๐ฅ105โค90
Runway Characters
Unlike HeyGen and Hedra, this is actually real-time โ you can have a conversation with them live (check out the examples).
And this is the first example of this level of quality in real time.
Itโs also a very direct response to a huge market demand: give our chatbots a face โ we want someone visible talking to the customer.
Itโs clear why itโs API-only โ everything runs on their servers, and Iโm guessing the price will be huge.
But the quality for real-time is ๐ฅ.
Unlike HeyGen and Hedra, this is actually real-time โ you can have a conversation with them live (check out the examples).
And this is the first example of this level of quality in real time.
Itโs also a very direct response to a huge market demand: give our chatbots a face โ we want someone visible talking to the customer.
Itโs clear why itโs API-only โ everything runs on their servers, and Iโm guessing the price will be huge.
But the quality for real-time is ๐ฅ.
๐2.21K๐30โค28๐ฅ22
The new Rotate Object feature in the latest Photoshop beta.
It looks good โ considering the input is just a flat image.
Then you can move on to lighting using Harmonize.
It looks good โ considering the input is just a flat image.
Then you can move on to lighting using Harmonize.
๐1.49K๐202โค194๐ฅ188
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
In Grok you can now add up to 7 references.
All of them appear in the video with normal consistency.
All of them appear in the video with normal consistency.
๐1.93K๐290๐ฅ280โค249
Hitem 3D has been updated to version V2.
This portrait generation from Nanabanana into 3D already looks quite appealing. See the second video.
One of the new features is Photo to STL for printing models on a 3D printer โ itโs essentially a packaged pipeline that combines generation and export to STL.
Itโs probably the most highly detailed 3D generator right now.
It can create models with up to 2,000,000 polygons.
https://www.hitem3d.ai/
This portrait generation from Nanabanana into 3D already looks quite appealing. See the second video.
One of the new features is Photo to STL for printing models on a 3D printer โ itโs essentially a packaged pipeline that combines generation and export to STL.
Itโs probably the most highly detailed 3D generator right now.
It can create models with up to 2,000,000 polygons.
https://www.hitem3d.ai/
๐1.13Kโค414๐398๐ฅ374
ByteDance, the parent company of TikTok, has postponed the global launch of Seedance 2.0. The reason is copyright claims from major Hollywood studios and streaming platforms. Reuters, citing The Information, reports that the launchโoriginally planned for the end of Marchโhas been put on hold, while the company has not yet publicly provided a detailed confirmation of these reports.
๐134โค43๐ฅ38๐30
Midjourney V8
The V8 alpha has been released. In short, itโs undertrained, still struggles with fingers, thereโs no editing yet, and the video model has been postponed (according to the developers).
I was going to write about the new features and how Midjourney is falling behind, but I found a great tweet that sums it up perfectly and matches my thoughts.
๐จ Midjourney launched the V8 alpha on March 17, 2026. http://alpha.midjourney.com is positioned as the fastest model yet, with better prompt understanding, improved text rendering, support for moodboards, sref, multiple aspect ratios, native HD 2K output, and image generation about 4โ5x faster than previous versions.
๐ค The problem is that, based on my tests, these promises donโt hold up. Incorrect hands, distorted proportions, weak anatomy, and overall quality that often feels worse than V7. What stands out most is that in a rapidly evolving AI image market, releasing such a raw alpha risks looking more like a response to competition than real progress.
The V8 alpha has been released. In short, itโs undertrained, still struggles with fingers, thereโs no editing yet, and the video model has been postponed (according to the developers).
I was going to write about the new features and how Midjourney is falling behind, but I found a great tweet that sums it up perfectly and matches my thoughts.
๐จ Midjourney launched the V8 alpha on March 17, 2026. http://alpha.midjourney.com is positioned as the fastest model yet, with better prompt understanding, improved text rendering, support for moodboards, sref, multiple aspect ratios, native HD 2K output, and image generation about 4โ5x faster than previous versions.
๐ค The problem is that, based on my tests, these promises donโt hold up. Incorrect hands, distorted proportions, weak anatomy, and overall quality that often feels worse than V7. What stands out most is that in a rapidly evolving AI image market, releasing such a raw alpha risks looking more like a response to competition than real progress.
โค64๐50๐ฅ48๐47
This media is not supported in your browser
VIEW IN TELEGRAM
Magnific Precision is a new video upscaler.
On Freepik, in the Video Upscale section, theyโve released a new model โ Precision.
While the previous model, Magnific Creative Video, was more geared toward NPR content (anime, 3D, cartoons, stop motion), Precision is positioned specifically as an upscaler for photorealistic footage (and all your generations from Siden and Kling).
You simply upload a video, set the resolution, and adjust a single parameter โ Strength, which controls how much detail is added.
Notable details:
โข Maximum of 7200 frames and 600MB
โข Up to 5 minutes at 24 fps
โข Supports 4K
โข Generates a half-second preview before processing the full video (convenient)
Thereโs a suspicion it might struggle a bit with highly detailed or noisy textures.
Letโs try it out.
https://www.freepik.com/ai/video-upscaler
On Freepik, in the Video Upscale section, theyโve released a new model โ Precision.
While the previous model, Magnific Creative Video, was more geared toward NPR content (anime, 3D, cartoons, stop motion), Precision is positioned specifically as an upscaler for photorealistic footage (and all your generations from Siden and Kling).
You simply upload a video, set the resolution, and adjust a single parameter โ Strength, which controls how much detail is added.
Notable details:
โข Maximum of 7200 frames and 600MB
โข Up to 5 minutes at 24 fps
โข Supports 4K
โข Generates a half-second preview before processing the full video (convenient)
Thereโs a suspicion it might struggle a bit with highly detailed or noisy textures.
Letโs try it out.
https://www.freepik.com/ai/video-upscaler
๐351๐121๐ฅ111โค84