Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients

Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients

Anthropic launched the Claude 4.5 AI model on Monday, stating the latest iteration can write code for longer periods of uninterrupted time and better process finance and science tasks, as the startup expands into business AI.

Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients
Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients

The Alphabet and Amazon.com-sponsored AI company is beating competitors to develop models capable of consistently running software and executing multi-step tasks, important for AI agents, that execute tasks on behalf of humans.

The Sonnet 4.5 model built a web app from scratch in internal testing, and one customer let the AI chatbot code for 30 hours, up from a seven-hour run produced by Anthropic’s previous Claude Opus 4 for another client, Chief Product Officer Mike Krieger explained.

Anthropic is aiming at power users and business clients instead of pursuing a viral consumer moment, he added.

Claude 4.5 is also more adept at finance and scientific reasoning and more proficient in computer use, answering around 60% on a test of operating-system flexibility compared to around 40% for earlier versions, the firm stated.

“It’s much more intuitive when you just watch the model interacting with a computer the way a human would if you’re not a programmer,” said Chief Science Officer Jared Kaplan.

Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients
Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients

In a separate announcement on Monday, Microsoft announced it would roll out new Microsoft 365 Copilot features based on Anthropic models such as “Agent Mode” in Excel and Word and an “Office Agent” for chat in Copilot, with PowerPoint to come.

Microsoft announced last week it would bring Anthropic’s models into Microsoft 365 Copilot to broaden beyond long-standing partner OpenAI.

Anthropic introduces Claude 4.5, promotes improved capabilities, aims at business clients

Anthropic, which was started by ex-OpenAI leaders, has prepared Claude for business use with guardrails it claims diminish unsafe outputs.

The firm has been selling Claude’s coding and data-analysis capabilities to regulated businesses and groups that need models to function on many different software tools.

Read More: WhatsApp launches new shortcut to chat with status viewers directly

Here are the top AI stories from the past week:

From AI behemoths announcing new U.S. data centers, to an intelligent airport corridor, this is AI Weekly.

Krieger explained that the company’s emphasis is on consistent, predictable performance for long tasks and not short demos.

Leave a Reply

Your email address will not be published. Required fields are marked *