Tavus just launched PAL Maker, a no-code platform that lets anyone build interactive AI video companions simply by describing them to a creator agent named Charlie. Powered by models like Raven-1 and Phoenix-4, the avatars come equipped with faces, voices, vision, and long-term memory. Why it matters: It lowers the barrier of digital human creation to a casual chat, already serving roles from post-surgery check-ins to candidate screening.
Researchers just introduced LeapTalk, a framework designed to conquer the latency-quality trade-off in talking-head generation. By using Brownian bridge transport and heterogeneous distillation into a single forward step, it outputs lip-synced video at up to 200 FPS from one reference photo and audio, completely avoiding identity drift across long streams.
Mirage just launched Avatar X, a new model that creates full digital twins from just 10 seconds of source footage. Unlike existing avatar tools, it preserves micro-expressions like laughing, yawning, and smirking, driving realistic face and body gestures via audio across both horizontal and vertical long-form video.
Adobe and Johns Hopkins researchers just introduced Wonder, a video world model that turns a single image or video into an interactive environment. Users can explore scenes at 16 FPS with 6-DOF camera controls, and the model maintains consistent geometry even after roaming around for over a minute.
Google just open-sourced GNM, a generative 3D head model that gives creators granular slider-level control over facial features, down to the eyeballs, teeth, and tongue. With over 600 independent identity and expression parameters under an Apache 2.0 license, it tackles the "same-face" look and hollow-mouth flaws common in AI-generated humans.
Netflix’s Eyeline Labs just unveiled ID-V2V, a framework that restyles video lighting and scenes using a single edited keyframe. Crucially, it locks down the actor's facial identity, gaze, and lip sync, automatically correcting misaligned input frames. It maxes out at 720p, and with an 80GB model weight, you will need serious local hardware to run it.
Black Forest Labs just teamed up with mimic robotics to turn its new FLUX 3 multimodal foundation model into FLUX-mimic, a video-action AI running on robots in Audi factories. By tapping into the physics FLUX learns during video generation, a lightweight decoder translates scene dynamics into physical manipulation in under 80ms on a single RTX 5090.
Black Forest Labs just launched FLUX 3 into early access, a multimodal foundation model that jointly trains on images, video, and audio using its Self-Flow architecture. It generates up to 20-second video clips complete with synchronized native sound and even adapts to robotic action prediction. Open-weight Dev access and standalone image generation are slated to follow in the coming weeks.
ABot-World just open-sourced a real-time world simulator that lets you continuously explore and steer an AI-generated environment with action controls. It runs on a single desktop RTX 5090 GPU, delivering 720p video at 16 FPS with 1.2-second latency. Basically, instead of short video clips, you get an endlessly unfolding world running entirely on local consumer hardware.
Researchers just open-sourced Homie, a video personalization model that keeps complex characters and objects consistent across frames. It accepts multi-view inputs and OCR layout maps to accurately position text and logos, outperforming earlier models in multi-subject fidelity.
Neill Blomkamp just debuted NIGHTBORNE, a 13-minute short film created with ByteDance's Seedance 2.0 video model, real concept artists, and the likenesses and voices of 32 real people. After two years of experimentation, the District 9 director says he's now aiming to tackle a full-length feature and has launched an AI banner, Barley Studios.
Director Christopher Nolan just called AI a transparent Trojan horse made of glass, saying everyone already knows danger is lurking inside. Nolan praised public skepticism—particularly from younger audiences who dismiss AI content as slop—arguing that questioning creators' motives is essential to avoid blind faith.
Browser-use just merged a new feature into browser-harness that records browser agent sessions and automatically edits them into Screen Studio-style clips. The agent captures per-action screenshots, hides sensitive credentials, and renders smooth zooms, synthetic cursors, and captions inside the browser without needing ffmpeg.
Runway just launched Agent 2.0, turning simple text prompts into complete marketing briefs and multi-platform campaign assets. It can ingest performance data from platforms like Meta and YouTube, then orchestrate models like Gen-4.5 to output auto-formatted variants for different markets. Why it matters: Runway is making a serious play for end-to-end marketing workflows, moving far beyond standalone video generation.
Google just rolled out Gemini Omni editing and personal avatars to Google Vids. Users can now fine-tune lighting or swap backgrounds using conversational prompts, or generate a personal digital clone from a selfie and voice clip to deliver video updates camera-free. The features are available to Google AI Pro, Ultra, and Workspace business subscribers, complete with invisible SynthID watermarks.
Wan Streamer just introduced version 0.3, upgrading real-time interactive avatars from static talking heads into dynamic characters that move and react. By splitting video into a persistent scene world and a moment-by-moment event stream, it executes spoken action cues like gesturing or posing in real time. The research is currently detailed in a paper.
Netflix just disclosed in its Q2 earnings report that roughly 300 titles have used generative AI in production this year, spanning pre-visualization to post-production. Co-CEO Ted Sarandos highlighted 17 minutes of AI-enhanced footage in the docuseries "The American Experiment," produced twice as fast at half the cost. While Hollywood remains wary of automation, Netflix insists the tech is meant to empower human creators rather than replace them.
OpenMOSS just open-sourced MOSS-VL-Realtime, an 11B parameter vision-language model tailored for live streaming video. Instead of waiting for an entire clip to load before answering, it processes incoming frames on the fly, allows users to interrupt with questions at any timestamp, and stays silent when visual evidence is insufficient.
George Lucas just voiced strong support for AI in cinema, saying the technology makes filmmaking much easier and comparing pushback to preferring horse carriages over cars. The Star Wars creator argued that progress is unstoppable and AI can even help verify authentic content. It is a bold vote of confidence from a cinema pioneer while Hollywood remains fiercely divided over generative tools.
Alibaba's Wan team just open-sourced Wan-Dancer-14B, a video model that turns audio tracks and a single image into rhythm-synced, minute-scale dance videos. Unlike tools requiring reference choreography videos, it plans long-range choreography directly from music context across genres like hip-hop and classical. Why it matters: Long-form music-to-motion generation just got a massive open-source leap.
