Human versions Claude Sonnet 4.5 in the latest offer for artificial intelligence agents and coding

The latest model of artificial intelligence in Antarbur spent 30 hours running itself to codify the chat app closer to recession or difference. According to the Arthur, he only stopped running when she completed the task.

The model, Claude Sonit 4.5, today, is announced today, and its ability to work independently for 30 hours in a row is a big jump forward. Before, the company’s OPUS 4 model topped the headlines in the May for its ability to work for seven hours.

It is an important step in the Battle of Anthropor to the corner on the market on both artificial intelligence agents and coding of artificial intelligence. The company is called Claude Sonit 4.5 “the best model in the world for real world agents, coding and computer use” and said it “leads the market to use computers”, referring to the use of the computer for the first time Almost a year ago. The new model is particularly skilled in areas such as cybersecurity, financial services and research, according to anthropologist. One of the beta tests, Canva, said that the new model helped “in the long-context-context-engineering tasks in our database to the features and research of the product.”

The anthropologist, Openai, Google, and other companies launch the additional updates and features that allow their technology to work as an assistant to consumers (search for topics, scheduling meetings, search for flights) and use for institutions and developers (create chipset floors, assist in coding, and analysis dates). The battle rises for attention and dependence almost every month, if not every week. A few days ago, Openai Pulse announced, The latest ChatGPT feature It is designed to be part of user procedures in the morning and related search topics.

Anthropor also said that the new model will be associated with other updates to help developers coding their artificial intelligence agents.

“We combine the launch of the model with access to virtual devices, memory, context management and multi -agent support,” the company wrote in a statement. “These beams are essentially the same building blocks that are the Power Claude symbol – allowing developers to build their developed agents.”

Diane Bin, head of the product management in anthropology, told, freedom In an interview that the model improvements in their computer uses surprised her even. Claude Sonnet 4.5 is more than three times the skill in moving in a browser and using a computer compared to human technology last October. Ben said that the team had received notes from the customers who have made early arrival-“GitHubs and Coursors of the World”-and spent last month to work extensively on the model.

Scott White, Pioneer Product for Claude.ai, freedom The new model works at the “Employee President’s level” and it can find a availability between multi -people calendars and a meeting date, take a look at the data panel and withdraw visions together, and write status updates based on individual meetings with its direct reports, and more.

Neither White or Pence after coding in elegance with the new model Matthew tried freedom Talk to them. But Ben said that she was using Claude Sonnet 4.5 to employ potentially anthropologist members.

“It was really really useful to have a continuous wave under the operation I use,” Ben said, “Do a deep search on the web, and reach such parameters for the personal files of the source of certain types of roles in my team.” Email. “