Claude 4.5 Released Globally with Advanced Coding and Document Features

Claude 4.5 is currently available worldwide via API. Pricing remains the same as Claude Sonnet 4: $3 per million input tokens and $15 per million output tokens. For developers, API access is provided using the model identifier “claude-sonnet-4-5”.

Among the key innovations are various improvements across the Claude lineup. Thus, Anthropic added the ability to execute code and create files directly during conversations in the Claude web interface and the corresponding applications. You can also generate tables, presentations, and documents without leaving the chat.

The conflict initiators urged other citizens to commit unlawful acts, involving women and minors. As a result, stones and bricks began to be hurled at border guards and service vehicles, the press service said.

Expanded capabilities and new projects

In addition, Anthropic launched a five-day research preview called Imagine with Claude for Max subscribers, which demonstrates how the model creates programs in real time. The company describes this as “a nice demonstration of what Claude Sonnet 4.5 can do” when paired with the appropriate infrastructure.

The Claude Code command-line development tool has also received several updates: checkpoints to save progress and revert to previous states were added, the terminal interface was updated, and a native extension for VS Code was released. The Claude API also gained a new context-editing feature and a memory tool for long-running agent tasks.

Their actions show signs of criminal offenses under Article 345 of the Criminal Code of Ukraine “Threat or violence against a law enforcement officer” and Article 194 “Intentional destruction or damage of property.” The respective notices have been sent to the National Police

In the modern AI landscape, companies continue to cautiously assess progress using standard benchmarks, as measuring other aspects often remains a challenging task. Many users increasingly turn to chatbots as universal assistants. In light of the latest reports about issues with some users who delve into fantastical scenarios, Anthropic’s statements about reducing manipulation, deception, and the pursuit of power in Claude Sonnet 4.5 appear more pronounced.

Ultimately, the development of such systems underscores the constant balance between usefulness and caution in using artificial intelligence for a wide range of tasks – from automating code to managing business processes.