Anthropic shut down its first code with a cloud conference today with the announcement of a new frontier AI system. The company is calling Cloud Oppus 4 the world’s best coding model. According to anthropic, Opus 4 is dramatically better in functions that require thousands of different stages to be completed, giving it the ability to work continuously for several hours at a time. Additionally, the new model can use several software tools in parallel, and it is more accurately better in the following instructions.
In combination, anthropic states that those capabilities make Opus 4 ideal to power the upcoming AI agents. For unfamiliar, agentic systems are AIS designed to plan and do complex functions without human supervision. They represent an important step towards the promise of Artificial General Intelligence (AGI). In a customer test, anthropic worked on Opus 4 to seven hours, or about a complete working day. It is an important milestone that wants to build the company for the type of agentic system.
Another reason Anthropic thinks that OPS 4 is ready to enable the manufacture of better AI agents, as the model is likely to use 65 percent less shortcuts or flaws while completing tasks. The company says that the system displays much better “memory capabilities”, especially when developers clouds provide local file access. To encourage Devas to try Opus 4, making anthropic Cloud codeIts AI coding agent is widely available. It also added new integration with visual studio code and jetbrances.
Even if you are not a coder, anthropic may be something for you. This is because with Ops 4, the company announced a new version of its Sonnet model. Like the cloud 3.7 sonnet before and before the Opus 4, the new system is a hybrid regioning model, which means that it can indicate almost immediately and can engage in extended thinking. As a user, it gives you a best of both world chatbott that is better to deal with complex problems when needed. It also includes many of the similar improvements found in Opus 4, including the ability to use the equipment in parallel and the ability to follow the instructions more confidently.
Sonnet 3.7 was so popular among users that anthropic launched a maximum plan in response, starting at $ 100 per month. The good news is that you will not need to pay anywhere as close to using Sonnet 4, as Anthropic is providing it to free users.
For those who want to use Sonnet 4 for a project, API pricing is at $ 15 for $ 3 per million input tokens and $ 15 output tokens. In particular, outside all normal places you will find models of anthropic, including Amazon Bedrock and Google Wartax AI, Microsoft Sonnet 4 is being a default model for the new coding agent that is introducing through the Github Copilot. Both Opus 4 and Sonnet 4 are available to use today.
Today’s announcement has already been made during a busy week in the AI ​​industry. On Tuesday, Google announced its I/O 2025 conference with other things, saying it was rolling AI mode for all search users in the US. A day later, Openi Said that it was spending $ 6.5 billion to buy Johnny Ive’s hardware startup.
