Cursor launches Claude Max: 200,000 words context + 200 tool calls, a revolutionary AI model

Claude Max has now officially landed on the Cursor platform, and this new AI model based on Claude 3.7 shows outstanding capabilities in handling large-scale code projects. It is equipped with the latest Claude 3.7 engine, supporting a 200,000-word super-large context, allowing developers to load the entire project code at one time, and the model performance will improve with the context increasing.

The maximum number of tool calls is increased to 200, allowing it to edit, analyze and optimize the entire code base in one go without step-by-step processing. The code comprehension ability has been greatly improved, and you can understand a large amount of code at once, deeply understand the code structure, reduce unnecessary tool calls, and improve work efficiency.

However, Claude Max adopts a billing mode by usage, with a charge of $0.05 per request (prompt) and a tool call of $0.05 per tool call. If you use the limit of all 200 tool calls, the single operation fee can reach $10, so it is not suitable for daily use of ordinary users.

Claude Max is mainly suitable for professional developers who need to deal with super-large-scale complex projects, advanced users who pursue high-precision code editing, and users who have a sufficient budget to experience top AI capabilities. Cursor's normal model is enough for more than 90% of daily code editing tasks. Claude Max is not included in the basic Pro package and requires additional "Billing Mode" to be enabled.