Anthropic Rushes to Launch Cost-Effective Opus 5.5 Ahead of IPO as the Large Model "Cost-Reduction War" Heats Up

Anthropic PBC is releasing a new artificial intelligence model designed to handle a wide range of enterprise tasks at lower cost. The startup is preparing for a highly anticipated Wall Street debut, and the move is part of its effort to maintain its lead over competitors. The company officially released Claude Opus 5.5 on Tuesday, saying the model performs comparably to the more capable Fable 5.1 on most work tasks, but at a significantly lower cost than the previous Opus 5 version. Anthropic said the new model generates output faster than Opus 5 and writes more clearly. The AI developer plans to launch new versions of Claude Sonnet and Haiku in the coming weeks to expand its 5.5-series model family. Anthropic is stepping up development of more advanced and more cost-effective AI software, competing for customers with OpenAI, Alphabet's Google, and numerous Chinese companies. The Claude developer is expected to launch an IPO as early as this fall, while rival OpenAI will wait until next year. Opus 5.5 is also the first new model released since Anthropic CEO Dario Amodei called on the industry to support slowing the pace of AI development. He said the company will implement new safety measures, including bringing in third-party evaluators. This comes after growing warnings both inside and outside Silicon Valley about the potentially catastrophic risks of AI. Anthropic said Opus 5.5 was tested before release by external organizations including METR and Frontier Design. The company also deployed safeguards similar to those for its more powerful Fable model, aimed at restricting higher-risk cybersecurity and biological uses.