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

    Musk says Grok 4.8 will finish training the week of September 14, 2026

    Musk describes Grok 4.8 as a “2.5T model” trained with a new C++ software stack, with reinforcement learning planned next.

    EM
    B(
    LA
    4 Sources, ,

    TLDR

    Elon Musk said on September 14, 2026, that Grok 4.8 would finish training that week and start reinforcement learning. He described it as a “2.5T model” trained with “our new C++ software stack.”

    Combined views

    3.4M

    4 Sources, first seen 17d ago

    18.6K likes

    Combined views

    3.4M

    4 Sources, first seen 17d ago

    18.6K likes
    17d ago
    first seen 17d ago
    1.2K comments
    1.2K saves
    1.3K reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    1.2K comments
    1.2K saves
    1.3K reposts
    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    4 Sources

    @elonmusk@techdevnotes Grok 4.8, which is a 2.5T model trained with our new C++ software stack, will finish training this week and start RL
    @scaling01RT @elonmusk: @techdevnotes Grok 4.8, which is a 2.5T model trained with our new C++ software stack, will finish training this week and sta…
    @beffjezos@elonmusk @techdevnotes can you RL it on CAD more
    @rohanpaul_aiGrok 4.8 update from Elon Musk.

    4 Sources

    @elonmusk@techdevnotes Grok 4.8, which is a 2.5T model trained with our new C++ software stack, will finish training this week and start RL
    @scaling01RT @elonmusk: @techdevnotes Grok 4.8, which is a 2.5T model trained with our new C++ software stack, will finish training this week and sta…
    @beffjezos@elonmusk @techdevnotes can you RL it on CAD more
    @rohanpaul_aiGrok 4.8 update from Elon Musk.