Anthropic has opened an important function in the artificial intelligence chat application Claude. Introduced last year Artifacts The feature is no longer limited to viewing content. With the new Beta update, users can start developing direct applications in Claude. This structure works with a system that produces code according to the recipes given in the chat window. The application can also be visually monitored in a separate window on the same screen.
When the user tells Claude how an application should work, the system produces the corresponding codes. With the ongoing requests in the conversation, these codes can be updated or rewritten. Moreover, the changes are immediately reflected in the application area. Thus, even people who only explain what they want to do can get a software prototype. Thanks to this method, the need for technical knowledge decreases significantly.
Among the applications developed by the first test users, vocabulary, writing support tool And Data Analysis Module examples such as. Those who establish more complex structures have designed assemblies that Claude processes with multiple calls. In such examples, Claude’s control can be provided among the components that manage multiple tasks in turn. As the codes are produced by Claude, the communication of these components with each other is also through Claude’s API infrastructure. Thus, the systems created can distribute tasks within themselves.
The API consumption of the developed applications is deducted from the account of the person who uses only
Anthropic encourages users not only by developing applications, but also by sharing these applications. When a project is operated by another user, this use is written to the person who uses it directly, not to the developer. This may provide advantage in resource use, especially for people who develop vehicles open to common use. The party that carries the burden of the application becomes the person who uses it. Thus, no additional cost is required for shared applications.
Developed applications are shown on the Artifacts tab on the left side of the Claude application. It is possible to view or start a new version by copying previously created projects. Users can develop ideas for their own applications by examining the applications of others. Thanks to the fork feature, it is also supported to copy an existing work and advance in a different way. This structure provides the opportunity to learn from examples, especially for beginners.
However, the system comes with some limitations. In the current version, applications cannot call external API and do not have long -term data storage support. In addition, features such as file upload or export to the application are not added. Despite all these restrictions, the area offered for basic prototyping operations is sufficient for most users. Developers can continue their further work with Claude’s external vehicles.
This step by Anthropic makes it possible to create visual prototypes within a chat -based system. This structure, where interfaces that work without writing code can be created, offers an accessible experience especially for users who meet productive artificial intelligence solutions for the first time. The transformation of text -based explanations into operating software expands the daily usage areas of artificial intelligence. This model brings the user not only to the monitoring, but to a direct production position.