Inside the Deal

Genmo’s Open-Source GenAI Model Aims to Power the Future of Video

Today, Genmo announced a $30M Series A round led by NEA. The funding marks a big milestone for co-founders Paras and Ajay Jain, two brothers on a mission to ‘put a tiny filmmaker in the pockets of a billion people’ with their open-source AI model for video generation. It’s an instantly resonant way to describe a massive ambition, evinced by a founding team who knows exactly how difficult it will be to accomplish.

A challenge worth tackling

Video has been the biggest opportunity at every platform shift, from the Internet to mobile and now AI. And it’s a big market that keeps getting bigger—perhaps not a surprise, Paras notes, given that 30-50% of the human brain’s cortex is devoted to visual signals. Video represents 82% of Internet traffic today, with 5x growth in total traffic in just the last five years. YouTube alone serves up about a billion hours of video to users every day on its platform.

One might even think that the video market is saturated, given its dominance across most social platforms today. Yet there is still an insatiable demand for more video content–both on the creation and consumption side. Consumers are seeking personalized and more relevant content, and everyone is clamoring for new tools that streamline production and enable creative expression with no technical experience. A huge unmet need persists, and the powerful capabilities of AI are ushering in a new wave of solutions.

But making a dent in the gigantic video market is a huge undertaking–it demands an approach that is both broadly accessible and highly ambitious. It requires a solution that is built from the ground up and capable of achieving massive scale at great speed. And it can only be accomplished by founders with deep technical expertise in both AI and video, who truly understand the market, and who are incredibly tenacious and resourceful. 

A founding team to meet the moment

These dynamics were well understood by Paras and Ajay, but they became increasingly convinced that it was the right time—and they were the right team—to bring high-quality, easy-to-use, AI-generated video to the world. They saw a powerful opportunity to leverage so many elements of their prior research and work across diffusion models, image generation, systems, and 3D generation. And they believed that creating an open-source model would enable them to make the biggest impact in a massive market that touches all of us.

In building Genmo, Paras and Ajay are applying the extensive expertise each has garnered through years of research and product development across deep learning and AI. Their PhD work at UC Berkeley was conducted alongside renowned technologists and entrepreneurs like Ion Stoica, Joey Gonzales, and Pieter Abbeel—all of whom now serve as advisors to Genmo and whom we at NEA have had the privilege of working with. Both worked on high-profile Google Brain projects. Ajay served as an author on the 2020 diffusion model paper that would become the foundational architecture for image, audio, and 3D AI models. And few rival Paras’ expertise in building scalable systems, which is an integral but often overlooked part of creative tooling within GenAI. The complementary nature of their skill sets makes them an extraordinarily apt fit for the challenge they’ve set out to tackle.

On the leading edge of a massive opportunity

Genmo has already built an incredibly compelling model, and they’ve done it with a fraction of the resources deployed by others. (And Mochi 1 is now freely available for personal and commercial use under the Apache 2.0 license!) Their decision to build an open-source product is especially exciting to us, given NEA’s longtime experience in open-source. We’ve helped businesses with thriving open-source communities achieve massive scale, including Databricks, MongoDB, Elastic, and many more. Paras and Ajay have no illusions—they know the road ahead is long and rife with obstacles, but they are fiercely committed to achieving their long-term vision—'a tiny filmmaker in the pockets of a billion people that will shape the next generation of video for businesses, creators, and consumers.

At NEA, we’ve been at the forefront of major tech transformations for nearly 50 years. We’ve learned that in each big shift, there are always moments where massive potential begins to crystallize into meaningful impact. Accelerating adoption and high-value use cases make it tangible, and palpable. We're rapidly approaching one of those inflection points with AI. And just like other platform shifts we've seen in the last few decades, video will play a starring role.

We believe Genmo is building a winning foundation model for AI-generated video, and we’re thrilled to partner with Paras, Ajay, and the rest of their incredibly talented team as they seek to power the future of video creation. We can’t wait for the world to experience their own filmmaker!