Researchers just introduced Seoul World Model (SWM), the first city-scale simulation grounded in a real metropolis. By using RAG to anchor video generation to real street-view photos, it produces stable, multi-kilometer trajectories without visual degradation—and even lets you spawn Godzilla into the skyline via text prompts.
Alibaba just open-sourced Fun-CineForge, a multimodal dubbing model that syncs speech directly to characters' lip movements and timestamps. By treating time as an explicit modality, it handles fast speaker switches and facial occlusions cleanly, turning automated film dubbing into a realistic studio workflow.
Texas A&M researchers just open-sourced Pulse of Motion, a lightweight AI designed to fix "chronometric hallucinations" in generated videos. Instead of guessing playback speed, the sub-2GB model calculates the true physical frame rate directly from visual dynamics and retimes the clip so movements actually obey gravity.
ByteDance just paused the global rollout of its Seedance 2.0 video generator after facing fierce pushback from Hollywood and U.S. senators. Lawmakers demanded the app be shut down immediately over unauthorized use of copyrighted characters and actor likenesses. ByteDance says it is working on stronger safeguards.
Retime GS just showed how to build smooth, full-body 3D animations from regular 2D video frames. Unlike conventional 4D reconstruction tools that glitch when frames drop, it uses continuous-time 3D representations and improved 4D Gaussian Splatting to seamlessly fill in missing motion.
NBCUniversal just unveiled major AI upgrades coming to the Peacock mobile app. Headlining the rollout is "Your Bravoverse," which uses a Synthesia-generated avatar of host Andy Cohen and AI agents to curate personalized, swipeable feeds across 5,000 hours of Bravo content. Peacock is also rolling out AI-driven real-time cropping for live NBA broadcasts in vertical 9:16 this spring.
OpenAI just updated its Video API with new capabilities powered by Sora 2. Developers can now generate clips up to 20 seconds, insert custom characters and objects, export in 16:9 or 9:16 aspect ratios, extend scenes, and run batch generation jobs. It gives developers significantly more control when building video products.
OpenAI is reportedly planning to integrate its Sora video generator directly into ChatGPT, according to The Information. The move mirrors last year's rollout of built-in image creation, aiming to win back users amid stiff competition from Claude, though it could also spike operating costs and deepen deepfake concerns.
Runway just launched Runway Labs, an internal incubator dedicated to developing experimental applications powered by generative video and General World Models. Led by co-founder Alejandro Matamala Ortiz, the team will prototype tools across Hollywood, gaming, healthcare, and education. It signals Runway's push to move past basic text-to-video prompts and turn world models into commercial software platforms.
Researchers just open-sourced ID Laura (ID-LoRA), a unified digital human model that dumps the clunky two-step workflow of cloning audio and then syncing lips. Built on LTX-2, it generates matching video, cloned voice, and realistic ambient sound in a single pass from just an image and reference audio. Running it locally is heavy, demanding at least 24GB of VRAM.
NVIDIA just rolled out major updates for ComfyUI, adding a simplified App View that hides complex node graphs behind a clean prompt box. Creators also get an RTX Video Super Resolution node delivering 30x faster 4K upscaling, alongside NVFP4 precision on RTX 50 Series GPUs that slashes VRAM use by 60%. Why it matters: Running production-grade AI video locally on consumer hardware is becoming dramatically more practical.
UCL scientists just figured out how to reconstruct 10-second video clips solely by reading single-neuron activity in a mouse's visual cortex. By factoring in physical states like pupil dilation and movement, their dynamic neural model produced high-quality reconstructions that get sharper with more tracked neurons. Why it matters: It pushes neural decoding beyond fuzzy fMRI scans toward cell-level movie playback.
MatAnyone 2 just dropped to strip video backgrounds cleanly without green screens or studio setups. The tiny 140MB open-source model handles fast motion and hair strands with precision, and it is free to try directly on Hugging Face.
Tavus just launched Raven-1, a multimodal perception system that reads tone, facial expressions, and body language alongside spoken words. Instead of sorting feelings into rigid boxes like happy or sad, it generates descriptive natural language with context updated every 300ms. Why it matters: It stops conversational AI from relying on flat text transcripts, helping agents catch sarcasm and hesitation in real time.
Researchers just open-sourced CubeComposer, a video model that converts standard single-camera video into full 360-degree panoramic footage with 4K enhancement. By breaking frames into six cube faces and using sparse attention, it fills in what was behind the lens far better than tools like Argus, though minor distortion remains.
Tencent researchers just introduced EffectMaker, a framework that clones visual effects from a reference video onto new subjects. It pairs a multimodal LLM with a diffusion transformer to adapt complex effects without per-effect training, backed by a massive 130k-video dataset.
Lightricks just launched LTX-2.3, upgrading its open-source audiovisual model with sharper textures, cleaner dialogue, and native 1080x1920 portrait video. Alongside it comes LTX Desktop, a free local video editor powered by the same engine. Why it matters: Full model weights and ComfyUI workflows are live on Hugging Face, free for companies making under $10M in annual revenue.
RealWonder just open-sourced a system that lets users apply virtual forces to a single image and watch objects move in real time. Powered by a physics simulation bridge and a 4-step diffusion generator, it hits 13.2 FPS at 480p on an H200 GPU. It brings structural physical understanding to AI video, with weights and code fully public.
Luma just introduced Luma Agents, a creative agent platform powered by its new Uni-1 unified multimodal model. The system handles end-to-end ad campaigns across text, images, video, and audio, using an iterative self-critique loop while orchestrating third-party models like Google's Veo 3 and ElevenLabs. It is now rolling out gradually via API.
Netflix just acquired InterPositive, the AI filmmaking startup founded by Ben Affleck in 2022. Instead of scraping existing films, the team trains smaller models on proprietary footage shot in a controlled soundstage to master lens physics, lighting, and editorial consistency. Affleck joins Netflix as Senior Advisor to ensure AI supports human judgment on set rather than replacing it.
