Genmo

Motion art created by Alper Durmaz with Genmo.

Coming soon: new capabilities in Genmo AI Replay. What words will you create?

Fully AI generated food commercial, made with Genmo AI in <1 hour. All clips used default settings and simple prompts like "tomatoes flying and colliding in midair", with Replay v0.2 text-to-video. Via Csaba Kissi

We're thrilled to welcome Nick Liu to the Genmo AI team. Nick joined us earlier this month in San Francisco. He has tremendous experience in AI infrastructure: Nick previously built LLM serving technology at Cohere, and worked on performance and perception at Roblox, Argo AI and Kodiak. Welcome Nick - and Sora!

One month ago, we at Genmo AI announced Replay v0.1: a photorealistic, buttery smooth text-to-video generative model that lets you create photorealistic videos from simple text prompts. Today, we’re announcing Replay v0.2: an improved model with better quality and more controllability. It's available now. What’s new? 1) Prompting with an image: 2) 3x longer AI video with up to 6 second long clips 3) 2.7x higher resolution Image-to-video: Transform an image into a video. Just drag and drop any image, and it will be used as a style reference to generate a matching video. This was a top-requested feature from our community, so we're delighted to open up access today. Higher resolution & improved quality: Video clips created with Replay v0.2 have over 2.7x the pixels of Replay v0.1 results. Resolution is now consistently greater than 2K, whereas v0.1 created in 720p-1080p. Longer videos: Generate up to 6 seconds of video, 3x longer than Replay v0.1's limit of 2 seconds. We've been working very hard to develop this new model over the last month. #multimodal is the future for foundation models, and I'm so excited to be embarking on this journey towards #aivideo.

Introducing Replay Batch Mode. Iterate faster on your video prompts. With Genmo AI's batch mode, users can simultaneously create multiple video for the same prompt with a single click. Our robust computing infrastructure will automatically scale to meet your demand.

Generative AI is notorious for its computational cost. Users suffer: how is it remotely acceptable to leave your users staring at a blank screen + progress bar? We're all scarred by memories of the infamous "spinning beachball of death", and in 2023, we can do better. Today, Genmo AI announced Replay video streaming. In case you missed it, Replay is our latest text-to-video generative model. Now, users see pixels within 10 seconds of writing a prompt, and full animated previews within 20-30 seconds. Genmo builds crazy GPU serving infrastructure that greatly improves UX. There's a lot going on under the hood to make this possible so we're super excited for you to try it. See Replay streaming in action below.

Replay from Genmo AI now suggests follow-up prompts after every video you create. Just like DALL-E 3, riff on ideas with our AI copilot. Suggestions are rolling out now - never have writer’s block again. It’s a small button in the UI, but under the hood, we use the same multimodal LLM behind Genmo Chat to perceive and rewrite prompts for our video model. 🧑🎨 “A close up of an astronaut in orbit viewing the aurora” ⤵️ ✨ “Close-up of a space traveler observing a black hole” ✨ “Futuristic astronaut surrounded by alien flora and fauna” ✨ “An astronaut walking on the lunar surface”

Latest interview from Joseph Gonzalez is out! We chatted with Ajay Jain, Ph.D, who's the co-founder of Genmo AI. The conversations touches on the history of diffusion models 🖼️, some awesome demos 🎥, and the tech behind Genmo 🤖. Check it out! https://lnkd.in/gmMm_Uzs