NVIDIA used its SIGGRAPH announcements to line up a wave of creative software behind Model Context Protocol (MCP), the open standard that lets an AI agent read an application
NVIDIA's Synthetic Video Detector Puts a Deepfake Score Inside Newsroom Workflows July 20, 2026 NVIDIA introduced the Synthetic Video Detector NIM microservice at SIGGRAPH, adding an AI-assisted signal that flags whether a video clip contains synthetic content. It is part of
Alibaba's Qwen team has opened preview access to Qwen3.8-Max , a 2.4-trillion-parameter model the company describes as its most capable system yet, with open weig
Google has open-sourced GNM Head , a parametric 3D statistical model of the human head that gives artists slider-level control over facial identity and expression. The release is th
Moonshot AI’s Kimi K3 is a 2.8-trillion-parameter mixture-of-experts model with a 1-million-token context window and native multimodal input. The important detail for produc
NVIDIA's ARDY Generates Controllable Human Motion in Real Time, and Releases the Code and Weights July 17, 2026 NVIDIA's Spatial Intelligence Lab published ARDY , a framework that generates 3D human motion in real time from text prompts a user can change while the charact
Developer Alex Barashkov has released Aval , an open source format for interactive video on the web that responds to hover, click, and application state instead of playing back as a fixed
Thinking Machines Lab released Inkling , its first in-house AI model, and put the full weights up for download. It is an open-weights, multimodal Mixture-of-Experts model that ta
Foundry has released SmartRoto, an AI rotoscoping add-on for Nuke that fills in the frames between an artist's keyframes. According to Foundry , the tool moves and deforms splines across a s
Mirelo AI and Kyutai open-sourced MuScriptor , an audio-to-MIDI model that takes a finished music mix and transcribes every instrument at once into separate MIDI tracks. Mos
Robbyant, an embodied AI company inside Ant Group, released LingBot-World 2.0 , an open-source real-time world model that builds an explorable environment from a single image. The model is liv
Shanda's Alaya Lab released AlayaWorld , an interactive world model that generates video environments you can steer in real time, and published it as a full-stack open-source framewo
Meta Superintelligence Labs launched Muse Image, an image model that can call its own coding and web-search tools mid-generation, and opened an early preview of Muse Video, a compan
Beeble's SwitchHDR Reconstructs True 16-Bit HDR From 8-Bit Video July 03, 2026 Beeble released SwitchHDR , a model that reconstructs true 16-bit HDR from ordinary 8-bit SDR footage. Instead of stretching the existing signal, it infers the high-dynamic-range scene, rebuilding
Seedance 2.0 Mini, a lighter variant of ByteDance's Seedance 2.0 video model, is now available , generating clips at roughly twice the speed and about half the cost of the standard model while h
Michael Caine's Licensed AI Voice Narrates ElevenLabs' 13-Hour 'Odyssey' Audiobook June 23, 2026 ElevenLabs has released a roughly 13-hour AI-generated audiobook of Homer's The Odyssey through its ElevenReader app, narrated by a licensed AI replica of Michael
ByteDance is advancing its Seedance video models on two tracks at once. Seedance 2.0 now generates 4K output, an upgrade the company confirmed at its Volcano Engine FORCE conference. Alongsid
Krea has released the open weights for Krea 2 , its in-house foundation image model, splitting the release into two checkpoints aimed at different jobs. Both are available to download now from
Sim-Plates.ai turns the company's CG driving-plate library into an AI generator that swaps lighting, weather, and era without touching camera motion June 22, 2026 Sim-Plates founder Alex Pearce has unveiled Sim-Plates.ai, an in-development AI plate-generation app that pairs t
HeyGen has introduced frame.md , a brand-spec format aimed at letting AI agents generate branded video and motion content that stays on-brand. The company is positioning it as a video-focused coun
Ideogram released Ideogram 4.0 with downloadable model weights, available on every Ideogram plan and through the company's API on launch day. The company announced the release on X, posit
Reve has launched Reve 2.0 , a new image model that the company is pitching on two claims: that it is "the best 4K image model in the world," and that it introduces "a new
Miso Labs has released the open weights for Miso One, an 8-billion-parameter text-to-speech model the company pitches as "the most emotive voice model in the world." Co-founder A
Google has introduced Gemini 3.5 Flash , positioning it as the first model in what the company calls its next generation of Gemini — and the performance backbone behind a faster, more
