Meta Launches Muse Spark Open to Rival Chinese Models
Meta has thrown its hat into the increasingly crowded open-weight model race with the release of Muse Spark Open, a sprawling 816-billion-parameter system built to handle text, images, video, and audio in a single unified model. The sheer scale of the release signals that Meta wants to compete not just on size but on versatility, aiming squarely at the kind of all-purpose capability that has made…
Continue reading →Need music for your AI video?
Epidemic Sound gives you unlimited access to 55,000+ tracks and 250,000+ sound effects — copyright-safe, so your videos never get flagged. Start with a free trial.
Latest
Tavus Avatar Fools Nearly Half of Callers in Test
Tavus says its new Griffin avatar passed as human to 48% of people on a one-minute video call, a dramatic jump from prior systems that barely registered a few percent.
Researchers Blend Real Podcast Footage With AI Faces
A research team has published demos of a system that splices authentic Joe Rogan Experience footage featuring Elon Musk with AI-generated synthetic faces and voices, alongside a write-up examining the risks such blended media poses.
PixVerse Launches R2 Real-Time World Model
PixVerse announced R2, a new real-time world model built for interactive, prompt-driven scenes that can maintain persistent characters as they unfold.
World Labs Tests Atlas Model On NVIDIA's Own HQ
World Labs ran its Atlas world model on Voyager, NVIDIA's headquarters, feeding it just 32 photos to see how well it could rebuild and fly through a real, familiar space.
Google Teases New Gemini Flash Text-to-Speech Models
An announcement attributed to a Google figure claims new Gemini Flash and Flash-Lite TTS models with thousands of voices, cloning, and broad language support, though the news comes solely from a social post so far.
Runway Explores Streaming AI Video Generation
Runway is researching a way to generate AI video as a continuous live stream that users can steer in real time, rather than waiting for finished clips to render.
OpenAI Begins Rolling Out GPT-6 Sol and Luna
OpenAI has started deploying two new GPT-6 variants, Sol and Luna, across ChatGPT, Codex, and its API, with paid tiers getting full access first and free users getting a limited preview.
Startup Claims Neuron-Derived Chip Speeds Up AI Video
The Biological Computing Co. says a tiny software add-on built from lab-grown neuron activity could make text-to-video generation five times faster and far cheaper, though the claims remain unverified and the product isn't public yet.
Google DeepMind's Lyria 3.5 Launches on fal
fal has added Google DeepMind's Lyria 3.5 text-to-music model to its platform, letting users generate songs from written prompts.
We Analyzed 15+ AI Video Models: Here Are the Cheapest Per Second
A breakdown of 15+ top AI video models reveals that generation costs range from under $0.02 to $0.40 per second, highlighting massive price differences between budget tiers and premium flagship engines.
Nvidia Breaks Down Mixture-of-Experts Model Efficiency
Nvidia published a technical explainer contrasting dense and mixture-of-experts architectures, showing how a 30-billion-parameter model can run using only 3 billion active parameters per token.
Google Debuts Gemini 3.8 Live Audio Models
Google has introduced two new real-time voice models, Gemini 3.8 Live and a heavier 'Extended Thinking' variant, promising top-tier performance at competitive pricing along with broad multilingual support.
GPT-6 Astra Still Can't Natively Generate Video
A capability breakdown of GPT-6 Astra shows the model remains a text-in, text-out system that can call an image tool, but has no built-in support for audio or video generation.
NVIDIA and USC Tackle Video Relighting Flicker
A new research effort called HorizonRelight aims to fix the visible lighting jumps that appear when diffusion models relight long videos in short chunks.
ElevenLabs Rolls Out Music v2.5 Scoring Tools
ElevenLabs is pushing its updated Music v2.5 model into its ElevenCreative platform and developer API, aimed at composing soundtracks for ads and branded video content.
ElevenLabs Upgrades AI Music Tool to Version 2.5
ElevenLabs has rolled out Music v2.5, an update to its ElevenMusic generator that promises fuller melodies, more realistic instrument sounds, and richer arrangements aimed at commercial users.
MODELSSep 10, 2026, 9:48 PMAvataar Builds Video AI Under India's AI Mission
Startup Avataar is developing a generative video model tailored to Indian languages and culture, one of a dozen companies tapped by India's government-backed AI Mission for subsidized compute.
OpenAI Rolls Out ChatGPT Images 2.5 Update
OpenAI has begun rolling out ChatGPT Images 2.5 to all ChatGPT, ChatGPT Work, and Codex users, alongside two new image-generation models for its API.
Runway says it's doubling down on real-time video generation
Runway says it's doubling down on real-time video generation, pointing to last week's Solaris and GWM Worlds 2 releases as steps toward instant generation becoming the industry norm.