
Models
New language, vision and audio models.
Page 4


Mistral releases weights for its trillion-parameter model, targeting cybersecurity
ModelsLaunch

Google's Gemini 4 Carbon tops AI rankings, but only cyber defenders get it

Gemini 4 Carbon tops the Artificial Analysis Intelligence Index

Mistral says Le Chonk has leapfrogged Chinese models in the open-weight race
ModelsLaunch

Mistral releases open weights for its 1 trillion-parameter Large 4 model
ModelsLaunch

Report: DeepSeek and Moonshot trained new models on GPT-6 Astra

Block unveils Dessa, a full-duplex speech model trained only on speech
ModelsLaunch

Google announces Gemini 4 Carbon with 'significant gains across coding and agentic tasks'
ModelsLaunch

NVIDIA unveils Nemotron 3.5 Transcribe, a promptable multilingual speech recognition model
ModelsLaunch

Nvidia releases open transcription model supporting up to 40 languages
ModelsLaunch

Liquid AI unveils LFM3 for specialized multimodal models
ModelsLaunch

Google DeepMind unveils Genie 4 world model with audio, realistic worlds and real-time steering
ModelsLaunch

Google shares eight examples of Gemini 4 Argon in action

Google's Gemini 4 Argon model goes live in the Gemini app

OpenAI adds GPT-6.1 Terra model to ChatGPT and Codex
ModelsLaunch

Hugging Face releases Carbon-X, a 14B open model that brings DNA and natural language together
ModelsLaunch

Block unveils Dessa, an AI model trained only on human conversations
ModelsLaunch

NVIDIA releases Nemotron 3.5 Bolt E4B, an open model for local voice agents
ModelsLaunch