Anthropic on Monday launched a new artificial intelligence (AI) model called Sonnet 5.5, the company's second model release since CEO Dario Amodei called for slowing down the pace of AI development.

Sonnet 5.5 is a faster and cheaper product from Anthropic. The model went live Monday across all platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure. Sonnet 5.5 is priced at $2 per million input tokens and $10 per million output tokens, which is half the price of Opus 5.5.

Anthropic said that compared to the previous generation, Sonnet 5.5 also requires fewer tokens to complete each task, making it cheaper to run. Anthropic stated that compared to its predecessor, Sonnet 5.5 performs better at coding, completing well-scoped tasks, and producing polished documents, slides, and spreadsheets.

The release comes less than a week after Anthropic introduced the more expensive Opus 5.5. Anthropic also said Monday that Haiku 5.5, the lowest-priced model in its product lineup, is coming soon. Theo Chu, a research product manager at Anthropic, said in an interview: "Sonnet is really aimed at cost-conscious customers who may not need that level of intelligence. Some tasks may just require performing routine work and don't need the kind of judgment that Opus can provide."

Over the past few weeks, as more researchers have warned about the potentially catastrophic dangers of AI, Anthropic and its competitors have come under close scrutiny over their safety and security measures. Anthropic CEO Dario Amodei called for slowing the pace of frontier AI model development in a warning piece published on September 12. In the article, he stated bluntly that the risks posed by AI are "severe" and that time must be taken to address them. His call was quickly echoed by two key figures, Musk and Altman.

Anthropic said Monday that Sonnet 5.5 did not push the frontier of its model capabilities, and therefore, according to a company blog post, most of the model's alignment testing focused on "a specific set of risks that apply to models at any capability level." In the AI industry, an "aligned" model refers to one whose behavior is consistent with human interests and values.

However, Anthropic said Sonnet 5.5's cybersecurity capabilities have been "substantially improved" compared to the previous generation, making it the first Sonnet model to launch with fallback mechanisms and cybersecurity safeguards similar to those Anthropic developed for its most capable models. Theo Chu said: "From the very beginning, prioritizing alignment and safety has been a key part of our mission. This has always been a priority for us across all models."