A stealthy AI image generator for branding has emerged from the shadows with a major funding boost. Recraft, the San Francisco-based startup behind a model that quietly outperformed OpenAI’s DALL-E and Midjourney on a leading industry benchmark, has secured a $30M Series B round led by Accel. The company raised $12M from Khosla Ventures in 2024 and now serves 4M users and has passed $5M in annual recurring revenue.
Recraft’s promise is clear for marketers drowning in last-minute campaign demands or designers struggling to generate brand-compliant imagery at scale. The company’s proprietary AI image generator, explicitly designed for branding use cases, lets teams create polished visuals with precise logo placement and strict adherence to brand guidelines, without hours of manual tweaking.
The model that put Recraft on the map, dubbed “red panda,” quietly topped the respected Artificial Analysis benchmark last year. This marked a significant leap in generative AI performance, especially given that its creators had not revealed the technology’s existence before the win. According to founder and CEO Anna Veronika Dorogush, the model got its nickname because users kept generating pictures of red pandas during testing.
Meet Anna Veronika Dorogush: A founder breaking moulds.
Recraft’s journey is deeply tied to Dorogush’s unusual career path. As a solo female founder and CEO, she stands out in the male-dominated world of AI startups, not just for her gender, but for her remarkable blend of experiences. Before entering tech, Dorogush worked as a professional model while earning a mathematics and computer science degree at one of Russia’s top universities. This rare combination of creative and analytical pursuits would later inform her approach to building AI systems that blend technical excellence with innovative design needs.
After her studies, Dorogush built an impressive technical resume, gaining experience at Google and Microsoft before leading machine learning work at Russian tech giant Yandex. There, she created CatBoost, an open-source machine learning library that became one of the world’s most popular ML platforms and is now used by companies like Cloudflare and JetBrains.
Her modelling background, she says, left a lasting impression: “The biggest lesson from that time was that grinding isn’t everything. When building a company, I know that to succeed, we have to be excellent at what’s mission-critical. In our case, building models is very important. So, we have put all the effort into excelling at this.”
Dorogush’s ability to merge technical depth with creative application has shaped Recraft’s distinctive positioning in the competitive generative AI landscape. Her leadership and vision have helped the company quietly outpace industry giants, proving that diverse backgrounds can lead to breakthrough innovation.
Why this AI image generator stands apart
Unlike most players in the generative image space, Recraft builds its models entirely from scratch. The company directly competes with names like Midjourney, Adobe Firefly, Stable Diffusion, and Black Forest Labs. However, its core edge lies in offering tools tailored to brand-focused output.
Recraft’s model lets customers place logos exactly where needed or easily create brochures and posters that follow a brand’s existing design language. This precise control is often lacking in other leading image models. It’s also why Recraft increasingly finds itself positioned closer to design tools like Canva, which recently added an AI generator of its own.
What makes Recraft V3 exceptional isn’t just its top benchmark ranking but its unique capabilities, specifically designed for professional designers and marketers:
- It maintains style consistency across multiple images, solving a major pain point for brand-focused design teams.
- It secured the #1 spot on the Artificial Analysis leaderboard with an impressive ELO rating of 1172, significantly ahead of competitors.
- It generates images in a median of 7 seconds, roughly twice as fast as OpenAI’s DALL-E 3.
- It’s the first model to offer image generation of any size and precise text positioning within images.
- It provides enhanced precision for branding elements, allowing logos and brand assets to be placed precisely where needed.
A high-profile investor lineup
The $30M Series B round was led by Accel, with continued participation from Khosla Ventures and new backing from Madrona. Khosla Ventures, a prominent deep tech investor, is known for early bets on transformative AI companies, including OpenAI and Anthropic. Accel’s involvement underscores confidence in Recraft’s commercial potential, while Madrona brings strategic experience from funding enterprise-focused startups across cloud, machine learning, and design tools.
Recraft’s total funding is $42M combined with its previous Series A. The company plans to channel the new capital into product development and expanding its user base, particularly among brands and agencies seeking fast, high-quality creative content generation.
Recraft hopes to carve out a durable position in the crowded AI image generation market by targeting a segment where visual accuracy and brand consistency are non-negotiable. For design and marketing teams facing tight deadlines and brand consistency requirements, Recraft’s technology offers a transformative solution beyond what general-purpose AI image generators provide.
