model
-
Blog
Google launches new genAI model for robots – Computerworld
Google subsidiary Deepmind has unveiled Gemini Robotics On-Device, a new version of the Gemini AI model meant to be used in robots and work without an internet connection. The new model reportedly supports natural language, making it easy to control the robot’s movements. In terms of performance, Gemini Robotics On-Device performs almost as well as the connected Gemini Robotics, Techcrunch…
Read More » -
Blog
Google rolls out text-to-image model Imagen 4 for free
Google Imagen 4, which is the company’s state-of-the-art text-to-image model, is rolling out for free, but only on AI Studio. In a blog post, Google announced the rollout of the new Imagen 4 model, but reminded users that it’s free for a “limited time” only. Unlike the old text-to-image model, Imagen 4 offers significant improvements and takes the text-to-image generation quality…
Read More » -
Blog
Microsoft’s new genAI model to power agents in Windows 11 – Computerworld
“By separating the input tokens from output tokens, Mu’s one-time encoding greatly reduces computation and memory overhead,” Pradeep said. The encoder–decoder approach was significantly faster than LLMs such as Microsoft’s Phi-3.5, which is a decoder-only model. “When comparing Mu to a similarly fine-tuned Phi-3.5-mini, we found that Mu is nearly comparable in performance despite being one-tenth of the size,” Pradeep…
Read More » -
Blog
A reasoning model to rival GPT-4 at 0.5% the cost – Computerworld
“The entire reinforcement learning phase used only 512 H800s for three weeks, with a rental cost of just $534,700,” the company explained. “This is an order of magnitude less than initially anticipated.” However, industry analysts urge caution. “MiniMax’s debut reasoning model, M1, has generated justified excitement with its claim of reducing computational demands by up to 70% compared to peers…
Read More » -
Blog
Google’s Veo 3 video model is now live in Canva — here’s how to try it
Google’s latest AI video model, Veo 3 (that’s the one with audio) is now available inside Canva — giving creators of all levels a simple way to generate high-quality video clips from a text prompt. The new feature, called Create a Video Clip, is now live for Canva Pro ($15/month or $120/year per person), Teams, Enterprise and Nonprofit users. The…
Read More » -
Blog
OpenAI Launches o3-pro, Its Most Capable Reasoning Model Yet
On Wednesday, OpenAI released o3-pro, its most capable reasoning model on ChatGPT. In April 2025, the company released the standalone o3 model along with o4-mini. The o3-pro model uses the same o3 as the underlying model, but runs in a high-compute mode, which uses more computing power and extended thinking time to solve harder problems. OpenAI says o3-pro offers better…
Read More » -
Blog
How to Make Sense of Garmin’s Forerunner Model Numbers
Garmin’s Forerunner line contains some of my favorite watches—the 265, which is arguably still the best of the bunch; the 570, which newly replaced it and has some snazzy new features, and the 955, which is a heck of a deal for something so feature-rich despite its age. If these numbers mean nothing to you, read on—they’re actually pretty easy…
Read More » -
Blog
ChatGPT prepares o3-pro model for $200 Pro subscribers
OpenAI is planning to ship an update to ChatGPT that will turn on the new o3 Pro model, which has more compute to think harder. ChatGPT currently offers o4-mini, o4-mini-high, GPT-4.5, GPT-4.1, GPT-4.1-mini, and o3 for $20 Plus subscribers. ChatGPT $200 Pro subscription has all of those plus an o1 Pro model, which offers more compute to think harder. o1…
Read More » -
Blog
DeepSeek releases new version of its R1 reasoning AI model – Computerworld
Chinese AI startup DeepSeek has released an update to the R1 reasoning AI model that took the tech world by storm when it was launched at the beginning of the year. The open-source model sent shockwaves through the AI industry as its efficient use of compute and memory resources helped it match leading US models’ performance at a fraction of…
Read More » -
Blog
Mistral AI launches code embedding model, claims edge over OpenAI and Cohere – Computerworld
Codestral Embed is designed for use cases such as code completion, editing, or explanation tasks. It can also be applied in semantic search, duplicate detection, and repository-level analytics across large-scale codebases, the company said. “Codestral Embed supports unsupervised grouping of code based on functionality or structure,” Mistral AI added. “This is useful for analyzing repository composition, identifying emergent architecture patterns,…
Read More »