• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
  • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Reka AI and NVIDIA Release Real-Time Video Model

    Official announcement shares details on a steerable video model developed with NVIDIA and offered in closed beta.

    AV
    RE
    JB
    4 Sources, 27d ago, first seen 27d ago

    TLDR

    Reka AI posted on X that it built a real-time video generation model with NVIDIA. The 30B-parameter transformer produces 720p video at 24 fps and accepts natural language instructions to steer output mid-stream. The company states the model runs 11.8x faster on a single H100 with no major quality drop in its blind evaluation. The post frames the release as progress toward World Language Action Models that act inside generated scenes. A linked company page offers access through closed beta.

    Combined views

    12.6K

    4 Sources, first seen 27d ago

    Combined views

    12.6K

    4 Sources, first seen 27d ago

    108 likes
    Today's Rank

    β€”

    Not ranked yet

    Today's Rank

    β€”

    Not ranked yet

    108 likes
    5 comments
    36 saves
    26 reposts

    Sentiment

    Positiveβ€”β€”Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    5 comments
    36 saves
    26 reposts

    Sentiment

    Positiveβ€”β€”Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    4 Sources

    @RekaAILabsTogether with @nvidia , we've built a real-time version of our video generation model: 30B parameters, steerable mid-stream in natural language, 720p at 24fps. 11.8x faster on a single H100, no major quality drop in our blind eval. We're one step closer to World Language Action Models: AI that doesn't just generate the physical world, it acts within it. Closed beta open now, contact us for access. https://reka.ai/labs/research/real-time-video-generation #RekaLabs #NVIDIA #AIResearch #VideoGeneration #WorldModels
    @julbernerExcited to collaborate with @RekaAILabs on real-time steerable video generation using FastGen! 🎬 Check out our FastGen repo - we will also drop new models and methods soon: https://github.com/NVlabs/fastgen
    @ArashVahdatRT @julberner: Excited to collaborate with @RekaAILabs on real-time steerable video generation using FastGen! 🎬 Check out our FastGen repo…

    4 Sources

    @RekaAILabsTogether with @nvidia , we've built a real-time version of our video generation model: 30B parameters, steerable mid-stream in natural language, 720p at 24fps. 11.8x faster on a single H100, no major quality drop in our blind eval. We're one step closer to World Language Action Models: AI that doesn't just generate the physical world, it acts within it. Closed beta open now, contact us for access. https://reka.ai/labs/research/real-time-video-generation #RekaLabs #NVIDIA #AIResearch #VideoGeneration #WorldModels
    @julbernerExcited to collaborate with @RekaAILabs on real-time steerable video generation using FastGen! 🎬 Check out our FastGen repo - we will also drop new models and methods soon: https://github.com/NVlabs/fastgen
    @ArashVahdatRT @julberner: Excited to collaborate with @RekaAILabs on real-time steerable video generation using FastGen! 🎬 Check out our FastGen repo…