Cheaper AI models often cost more to run than expensive ones, new research finds
New research shows that in over 1 in 5 comparisons, the cheaper AI model actually costs more to run. The culprit is a cost that most providers do not clearly surface.
News
-
-
Base 44 now lets you clone any website in minutes
Base44 just released a feature that lets you paste any URL and instantly clone its UI into a functional app. It solves the “generic AI design” problem by letting users directly copy human-made websites.
-
Google announces Gemini 3.1 Flash Live to power AI agents
Google just released Gemini 3.1 Flash Live. It eliminates awkward pauses and is heavily optimized for developers building autonomous AI agents.
-
Zhipu AI drops surprise GLM-5.1 update to challenge Claude 4.6 in agentic coding
Zhipu AI just surprise-dropped GLM-5.1. It rivals Claude 4.6 in coding at 1/7th the price. Here’s how to test it today.
-
Notion 3.4 adds dashboards, presentation mode, and a redesigned sidebar
Notion 3.4 brings dashboards, a redesigned sidebar, presentation mode, and a tabs block that apparently existed internally for years before anyone admitted it.
-
Mistral launches Voxtral TTS, completing its open-source speech pipeline
Mistral launched a text-to-speech model that clones a voice from 3 seconds of audio and runs on 3GB of RAM.
-
Meta’s TRIBE v2 can predict how your brain responds to videos and sounds
Meta built a model that predicts how your brain responds to what you see and hear. They also own the world’s largest advertising platform.
-
Lovable adds AI penetration testing to its platform through Aikido partnership
Lovable just launched AI-powered penetration testing for $100. A traditional pentest costs up to $50,000. That gap deserves some scrutiny.
-
ComfyUI’s new Dynamic VRAM lets you run larger AI models with less RAM
RAM prices are up, AI models are getting larger, and your hardware is struggling to keep up. ComfyUI just shipped an update that helps with all three problems at once.
-
NVIDIA’s AutoGaze makes AI video understanding up to 19x faster by mimicking human attention
NVIDIA found a way to make AI video understanding up to 19x faster by teaching models to ignore the parts of a video that do not matter.
