Recently, Anthropic, a leading company in the field of artificial intelligence, officially released its latest "hybrid reasoning model" - Claude3.7Sonnet. The model is hailed by the company as the "smarest" AI product to date, aiming to solve more complex problems and show outstanding performance in multiple fields such as mathematics and coding. Anthropic said that the launch of Claude3.7Sonnet marks another major breakthrough in artificial intelligence technology in reasoning capabilities and application scenarios.
According to Anthropic, Claude3.7Sonnet is a unique "hybrid" model that combines the dual advantages of "inference mode" and traditional mode. In “inference mode”, the model can pause and think deeply about complex problems, providing more accurate answers; in traditional mode, it can generate responses in real time, ensuring efficient processing speed. This hybrid design allows Claude 3.7Sonnet to perform well when dealing with simple problems (such as “when is it now?”) and complex tasks (such as “planning a two-week trip to Italy and considering the weather in late March”).
In addition to the release of Claude3.7Sonnet, Anthropic has also launched a "limited research preview" of its "autonomous" programming tool Claude Code. The tool is positioned as an "active partner" for developers, and is able to search and read code, edit files, write and run tests, submit code to GitHub, and use command-line tools. The launch of Claude Code further expands the application scenarios of the Claude series models and provides developers with stronger technical support.
Starting this Monday, users can already use Claude 3.7Sonnet in the Claude application, and developers can also obtain the model through Anthropic's API, Amazon Bedrock and Google Cloud's Vertix AI. It is worth mentioning that the operating cost of Claude 3.7Sonnet is consistent with the previous generation of 3.5Sonnet, with a charge of US$3 per million token input and US$15 per million token output. In addition, the model's knowledge deadline is October 2024, which is more recent than other models.
Anthropic also provides developers with a unique feature - the "draft paper" function, allowing developers to guide the model's "thinking" way, and even accurately control the model's response time. This function provides developers with more flexibility and control, allowing Claude3.7Sonnet to better adapt to various complex application scenarios.
In general, the release of Claude3.7Sonnet not only demonstrates Anthropic's technical strength in the field of artificial intelligence, but also provides users and developers with smarter and more efficient solutions. With the continuous development of AI technology, Claude3.7Sonnet is expected to realize its potential in more fields and promote the further popularization and deepening of artificial intelligence applications.