TEMPO.CO, Jakarta – Anthropic on Tuesday launched Claude Opus 5.5, the first model in its new Claude 5.5 family and its first major model release since CEO Dario Amodei called for the pace of frontier AI development to be slowed.
Anthropic said Opus 5.5 delivers performance comparable to Claude Fable 5.1 on most tasks while costing 40 percent less to run than its predecessor, Claude Opus 5, according to Yahoo Finance report.
The new model is priced at $4 per million input tokens and $20 per million output tokens, compared with $5 and $25, respectively, for Opus 5. Anthropic said cache reads, which account for a large share of costs in agentic and coding workloads, will cost $0.20 per million tokens.
Tokens are units used by AI models to process text, generally representing words or parts of words.
Opus 5.5 Focuses on Performance and Efficiency
Anthropic said Opus 5.5 is its strongest Opus model to date, with improvements in coding, computer use and other complex professional tasks.
The company said the model generates responses more than 30 percent faster than Opus 5 and requires less computing power to serve. Anthropic also said typical workloads will cost about 40 percent less than with the previous model.
The company said early testers reported significant gains in complex work. In one test, Opus 5.5 completed a 680,000-line code migration in less than a day, while another test found it succeeded in improving the loading speed of a web application in 39 out of 40 attempts.
Opus 5.5 is available through Anthropic’s Claude platform and API, as well as Amazon Web Services, Google Cloud and Microsoft Foundry.
Anthropic Highlights AI Safety Measures
The launch comes as the AI industry faces growing concerns about the potential risks posed by increasingly capable systems.
Anthropic said Opus 5.5 underwent extensive alignment testing and external evaluations before its release. The company said it also includes safeguards for high-risk areas such as cybersecurity and biology.
According to Anthropic, Opus 5.5 performed better than its previous models in its automated behavioral audit, which tests AI systems across thousands of simulated scenarios. The company also said the model is more resistant to prompt-injection attacks and less likely to take actions outside its assigned boundaries.
Anthropic said it is continuing to improve its alignment research and develop new safety-training scenarios.
Launch Follows Amodei’s Call to Pace AI Development
The release is notable because it follows Amodei’s recent call for a more measured pace of frontier AI development.
Amodei argued that AI progress should be paced so that safety practices can stay ahead of increasingly capable models. Anthropic said the approach is intended to address potential risks while allowing society to benefit from AI advances, particularly in areas such as biology and medicine.
The debate has intensified within the AI industry.
Earlier this month, Anthropic researcher Jacob Coxon announced his resignation and criticized AI companies over what he described as inadequate attention to safety. Anthropic alignment science lead Evan Hubinger subsequently said he agreed with Coxon’s concerns and estimated that there was a greater than 10 percent chance advanced AI could cause human extinction.
Amodei later published his argument for slowing the pace of frontier AI development.
OpenAI CEO Sam Altman and SpaceX CEO Elon Musk have expressed agreement with aspects of Amodei’s position, while Nvidia CEO Jensen Huang has pushed back against predictions that AI could destroy humanity, saying such concerns are not grounded in science.
US President Donald Trump has also dismissed some warnings about AI as conspiracy theories.
Anthropic Plans More Claude 5.5 Models
Anthropic said Opus 5.5 is the first release in its Claude 5.5 family.
The company plans to release Claude Sonnet 5.5 and Claude Haiku 5.5 in the coming weeks, with the models expected to bring similar improvements in performance, efficiency and safety.
The launch underscores the challenge facing AI companies as they compete to develop more capable systems while addressing concerns over safety, security and the cost of running increasingly powerful models.
