HomeTagsMultimodal AI

Multimodal AI

Stories from Multimodal AI come first, newest first, then stories from connected topics (open weights, AI Models, Hugging Face, AI, Nvidia, AI Agents), each labelled with where it comes from. This view is not indexed by search engines.

Google Research Open-Sources StreetReaderAI for Blind Users

Google Research has open-sourced StreetReaderAI, a map app for non-visual street exploration, built with and for blind and low-vision users.

Just in

Latest

NVIDIA Releases Nemotron 3.5 Super VL, an Open Multimodal Agent Model

NVIDIA released Nemotron 3.5 Super VL, an open 120B multimodal MoE for agents. Benchmark and partner figures are NVIDIA and partner claims.

Prime Intellect Acquires Moondream; Founders to Lead Inference and Multimodal Work

Prime Intellect acquired Moondream. An internal evaluation comparing Moondream 3.1 with Meta's Muse Wonder is Prime's own unpublished work.

NEURA Robotics Unveils Neuron AI, a Model That Lets Robots Combine Sight, Touch, Sound and Memory

NEURA Robotics launched Neuron AI, a model combining vision, touch, sound and memory to run its robots. A laundry demo is planned for GTC Berlin.

Subquadratic Launches SubQ Snap, a Long-Context Decision Model for Enterprise AI Agents

Subquadratic launches SubQ Snap, a decision model for enterprise agents with up to 384K tokens of context, priced at $0.05 per million input tokens.

Fudan Spin-Out Mosi Intelligence Raises More Than 1 Billion Yuan in a Pre-A Round Led by Huakong Fund

Mosi Intelligence, incubated from Fudan's OpenMOSS team, has raised more than 1 billion yuan in a Pre-A round led by Huakong Fund. No valuation given.

AI Trails People on Visual Reasoning Test; Base Models Reason on Cue

Scale AI and Elorian's Humanity's Sixth Sense benchmark finds people score 93.1% on everyday visual questions while the best model, GPT-6 Astra, scores 53.6%, mostly because models misread the picture. A separate preprint, not yet peer reviewed, says base models can reason almost as well as RL-trained ones when their answer starts with the right word.

Reka’s Rho-1: One AI Model Trying to Do the Work of Many

Reka has released a research preview of Rho-1, a 19-billion-parameter model that it says can read and create text, images and video and output robot actions in a single network. The speed claims are the company's own tests.

NVIDIA Posts $115.8bn Q3 Revenue as Vera Rubin Ships

NVIDIA posted $115.8bn in third-quarter revenue, up 103% year on year, and said its Vera Rubin NVL72 platform has shipped.

Microsoft Unveils MAI Code 2, Cyber 2 Flash and Perception

Microsoft has launched MAI Code 2, Code 2 Flash and Cyber 2 Flash at Ignite, and said its Perception security system is now generally available.

OpenAI Launches GPT-6.2 Astra and Sol, Now on Bedrock

OpenAI has launched GPT-6.2 Astra and GPT-6.2 Sol in ChatGPT and the API, with same-day general availability on Amazon Bedrock.

Latest on TSN

Explore the TSN Atlas: Every topic we cover, mapped. Click a bubble to dive in.