Latest LLM releases

LLM Releases
  1. Moonshot releases Kimi K2.7 Code
    released - Moonshot AI source

    Open coding-focused Kimi model with 1T total / 32B active parameters, native image/video input, and always-on thinking mode.

  2. Z.ai releases GLM-5.2 with a 1M-token context
    released - Z.ai (Zhipu AI) source

    MIT-licensed GLM flagship focused on long-horizon coding, agentic engineering, and IndexShare sparse-attention reuse.

  3. MiniMax releases MiniMax-M3
    released - MiniMax source

    Native multimodal 428B/23B-active model with one-million-token context and MiniMax Sparse Attention.

  4. US government orders Anthropic to pull Fable 5 and Mythos 5
    withdrawn - Anthropic source

    Access suspended three days after launch under an export-control directive citing national security.

  5. GPT-5.6 announced with a 1.5M-token context window
    announced - OpenAI source

    OpenAI previewed GPT-5.6, claiming the largest context window of any frontier model.

  6. Claude Fable 5 released — first public Mythos-class model
    released - Anthropic source

    Available across the Claude API, AWS, and Microsoft Foundry.

  7. NVIDIA releases Nemotron 3 Ultra 550B-A55B
    released - NVIDIA source

    Largest Nemotron 3 model appears on NVIDIA NIM with downloadable weights, 1M context, and agentic reasoning positioning.

  8. Alibaba releases Qwen3.7-Plus multimodal agent model
    released - Alibaba (Qwen) source

    Lower-cost multimodal sibling of Qwen3.7-Max with text, image, and video input and a 1M-token context.