Mochi 1 by Genmo

« Generate 1080p 30fps videos up to 15 seconds long from text or images with consistent characters, cinematic camera movements, and realistic 3D physics. Features : multi-Shots, audio lip-sync,native audio, native-level text rendering, etc. »

« Generate 30-second native videos in a single pass, using up to 50 multimodal references (images, audio, videos) to ensure consistency in characters and scenes. Local edits are also more precise and don’t require regenerating the entire scene »

« An open-source video generator that combines text, images, video, and audio to produce 2K clips up to 15 seconds long, with native stereo sound generated in a single pass. API available »

« Google's multimodal video model that generates and edits clips based on a simple conversation. It allows you to extend scenes up to 40 seconds, control the start and end frames, and features native audio synchronization. Rendering up to 4K resolution »

« A single AI model to create realistic images, videos up to 20 seconds long, and native audio. Video generation includes native multilingual audio: dialogue, sound effects, and ambient sound are produced directly alongside the images, without the need for separate post-processing »

« Generate cinematic videos with synchronized native audio and more precise narrative control. Create realistic 6-second clips that can be extended up to 1 minute. Insert/delete elements and achieve smooth transitions via Flow »

« Create realistic 5s videos from text or images. Smooth movements, precise physics and stunning cinematic camera work! »

« Create cinematic and ultra-realistic videos with physical accuracy, temporal consistency, and total creative control. Ranked among the world's best video models with 1247 Elo points. Available for all styles: photorealism, animation, etc. »