The AI ​​AI model defeated Dall-E and Midjourney at a popular reference point. Its creator has just landed USD 30 million.

Recectft, a startup behind the mysterious picture model that Dall-E and Midjourney Openai defeated on a valued industry hill last 12 months, raised a round of $ 30 million of the B series under the leadership of ACCl, said only TechCrunch.

Other investors in this round are Khosla Ventures and Madron. With its registered office in San Francisco, Recotft previously collected a series A with a value of $ 12 million run by Khosl in 2024. Startup based in San Francisco claims that he recently spent $ 5 million in ARR and 4 million users.

The startup drew the attention of the industry when his model, defeated “Red_Panda”, was on the point of artificial evaluation reference last 12 months. It was in fact V3 Recoth, which deserved its name, because early users generated images of the charming mammal, founder and general director of Rectift Anna (*30*) Dorogush said TechCrunch.

Recraft claims that he is building his own models from scratch and compete with other painting generators, equivalent to Midjourney, Adobe Firefly, Stable Diffusion and Black Forest Labs, said Dorogush. But Ai Recote is particularly distinguished by generating images for brands. This means enabling them to position the logo exactly if needed without an additional editing or easy generation of recent marketing materials, equivalent to brochures and posters, which are in line with existing brand guidelines.

According to Dorogush, this is an area where existing painting models often fail. This brings compensate to competing with design tools equivalent to Canva, which also has You have a generator For branding purposes.

Rekraft is also noteworthy because having a solo founder and general director. Dorogush founded Rectift after years of working on machine learning at Yandex – a Russian Google competitor – along with previous work on Google and Microsoft.

Before building AI models, she worked as a skilled model, while she obtained the degree of mathematics and computer science at one of the best universities of Russia. She ended up leaving this work, but says that she taught her that they were just working hard – like showing infinite casting connections – it was not enough.

“The biggest lesson from that time was that grinding is not everything,” said Dorogush. “Now, when building a company, I know that in order to succeed, we must be perfect in the field of mission criticism. In our case, building models are very important. So we put all efforts in it.”