Anthropic’s Claude Opus 4 model can work autonomously for nearly a full workday

Anthropor started her first blog ever with Claude today’s conference with the announcement of a new border system AI. Claude Obus 4 calls on the world’s best coding model. According to Antarubor, OPUS 4 is very better in tasks that require this to complete thousands of separate steps, giving them the ability to work continuously for several hours in one. In addition, the new model can use multiple software tools in parallel, and better to follow the instructions more accurately.

Combated, Anthropor says these capabilities make OPUS 4 ideal for operating the upcoming artificial intelligence agents. For unfamiliar systems, agents are AIS designed for planning and carrying out complex tasks without human supervision. It represents an important step towards the promise of artificial general intelligence (AGI). In the customer test, the OPUS 4 anthropologist operates seven hours, or nearly a full working day. This is an important landmark of the type of delegate systems that the company wants to build.

Another reason Anthropor believes that OPUS 4 is ready to enable the creation of the best Amnesty International agents is that the model is less likely to use an acronym or vulnerability when completing the tasks by 65 percent. The company says the system also shows a much better “memory capabilities”, especially when developers give access to local files. To encourage Devs to the OPUS 4 experience, Anthropic creates the CLADE code, artificial intelligence coding agent, which is widely available. It has also added a new integration with Visual Studio Code and Jetbrains.

Even if you are not programmed, the Antarubor may have something for you. This is because besides OPUS 4, the company announced a new version of the Sonnet model. Like Claude 3.7 Sonnet before and OPUS 4, the new system is a hybrid thinking model, which means that it can implement claims almost immediately and participate in expanded thinking. As a user, this gives you the best in both worlds Chatbot, which is the best processing to address complex problems when needed. It also includes many of the same improvements in OPUS 4, including the ability to use tools in parallel and follow the instructions more honestly.

Sonnet 3.7 was very common among anthropron users in providing a MAX plan, which starts from $ 100 a month. The good news is that you will not need to pay anywhere near a lot to use Sonnet 4, because the anthropoor makes it available to free users.

For those who want to use Sonnet 4 for a project, API pricing is residing at $ 3 per million input symbols and $ 15 for the same exit icons. It is worth noting that outside all the usual places you will find anthropor models, including Amazon Bedrock and Google Vertex AI, Microsoft makes Sonnet 4 the default model for the new coding agent that it offers through GitHub Copilot. Both OPUS 4 and Sonnet 4 are available for use today.

Today’s advertisement comes during what was already a week crowded in making artificial intelligence. On Tuesday, Google started the I/O 2025 conference, announcing, among other things, that the artificial intelligence mode of all research users in the United States has been presented. A day later, Openai said it spent $ 6.5 billion to buy Johnny Ev to start operating devices.