Posts in AI （699）

OpenAI Releases GeneBench-Pro, a New Benchmark for AI Scientific Judgment

Google Brings Nano Banana 2 Lite Image Model and Gemini Omni Flash Video Model to Developers

NVIDIA Cuts Token Costs by Up to 5x With Its Inference Software Stack, in One Month on Blackwell

NVIDIA Releases Agent Skills to Speed Up Vision AI Agent Development, Boosting Accuracy With Synthetic Data and Fine-Tuning

Google Opens Personalized Image Generation in Gemini to U.S. Users for Free, Linking Google Photos and Nano Banana

NVIDIA Jetson to Run in Lunar Orbit for the First Time via Firefly Ocula

Gemini Now Takes Notes in Google Meet for AI Pro and Ultra Subscribers

Palantir Builds Secure AI for US Agencies on NVIDIA Nemotron, Letting Them Own Models In-House

* Image is for illustration purposes.

xAI Makes Grok Imagine Video 1.5 Generally Available, With Synced Audio and 720p Clips in About 25 Seconds

OpenAI Previews Next-Generation GPT-5.6 with a Three-Tier Sol, Terra, and Luna Lineup

* Image is for illustration purposes.

OpenAI Extends ChatGPT Long-Paste Attachments to Free and Go, Raising the Threshold to 10,000 Characters

* Image is for illustration purposes.

ByteDance Unveils Seedance 2.5: 30-Second AI Video in a Single Pass, Up to 50 Reference Inputs

* Image is for illustration purposes.

OpenAI Publishes Economic Research on Codex, Says 99.8% of Its Internal Output Tokens Now Run Through the Agent

Amazon-Owned Zoox Overhauls Its Robotaxi Ahead of Commercial Service, Refreshing Interior and Exterior

* Image is for illustration purposes.

Mistral Launches OCR 4: Structure-Aware Document AI Spanning 170 Languages with Self-Hosting

* Image is for illustration purposes.

Google Gemini 3.5 Pro Nears June Launch with 2M-Token Context and Deep Think

Google Adds Built-In Computer Use to Gemini 3.5 Flash for Agentic Automation

OpenAI and Broadcom Unveil Jalapeño, OpenAI's First Custom AI Inference Chip

* Image is for illustration purposes.

Perplexity Computer Comes to Microsoft 365, Working Directly in Word, Excel, PowerPoint, and Outlook

* Image is for illustration purposes.

Google Brings Gemini to Chrome on Android, Including "Auto Browse," Rolling Out Late June

* Image is for illustration purposes.

Google Sheets Adds One-Click Gemini Fix for Formula Errors with Plain-Language Explanations

* Image is for illustration purposes.

OpenAI Adds Record & Replay to Codex: Demonstrate a Task Once and It Repeats Automatically

* Image is for illustration purposes.

NVIDIA and AWS Expand AI Production Infrastructure with New EC2 G7 GPU Instances, Up to 4.6x Inference

* Image is for illustration purposes.

Midjourney Builds a 60-Second Full-Body Scanner, Moving from Image-Generation AI into Healthcare