Alice’s July update spans images and accessibility
Yandex’s monthly digest combines creative and everyday assistant features.
Gemini Robotics 2 extends intelligence across the whole body
Google focuses on movement, dexterity, and cooperation between robots.
Cosmos-H-Dreams generates simulations for surgical robotics
NVIDIA describes a faster way to explore action-conditioned surgical scenes.
North Automations connects isolated agents into workflows
Cohere introduces orchestration for multi-step enterprise processes.
Opus 5 targets stronger everyday professional work
Anthropic updates its flagship with an emphasis on efficiency.
Claude can query its own Economic Index
Anthropic makes its usage research easier to explore conversationally.
Google splits Flash into more specialized roles
Gemini 3.6 Flash, Flash-Lite, and Flash Cyber address different agent workloads.
Grabette records robot data without starting with a robot
Pollen Robotics proposes an open handheld system for manipulation demonstrations.
The University of Toronto brings AI into institutional workflows
Cohere’s partnership combines private deployment with education and research.
Inkling brings Thinking Machines into open multimodal models
Hugging Face details a release spanning text, images, and audio.
VoiceEQ looks at what speech benchmarks miss
Hume AI researchers examine the human quality of voice interaction.
GPT-5.6 offers several balances of capability and cost
OpenAI launches Sol, Terra, and Luna for different workloads.
GPT-Live makes voice a model family
OpenAI introduces a new generation of models for spoken interaction.
Robostral Navigate uses one camera to get around
Mistral introduces a model for following navigation instructions in physical spaces.
LeRobot 0.6 closes more of the learning loop
World models, reward models, and evaluation tools join the robotics toolkit.
Leanstral 1.5 makes formal proofs more accessible
Mistral releases an Apache-licensed model for Lean 4 proof engineering.