Anthropic launched its Claude Opus 4.6 artificial intelligence model on Thursday, representing an upgrade to the Opus 4.5 model released in November. The San Francisco-based technology startup said the new model can work on tasks for longer periods and more reliably than its predecessor.

The company reported performance improvements in coding and finance applications with the new model. Anthropic is backed by Amazon.com (NASDAQ: AMZN) and Alphabet's (NASDAQ: GOOGL) Google.

The Claude Opus 4.6 model can process 1 million tokens in a single prompt, matching capabilities previously announced by Google and available in a less powerful Claude model. Tokens represent pieces of data processed by AI systems.

Anthropic also previewed functionality for its computer programming tool Claude Code, which can distribute tasks among multiple autonomous agents to complete work more efficiently. The announcement follows recent product advances that contributed to a selloff in traditional software stocks.

The information is based on the company's statement regarding the model launch and its technical capabilities.