Anthropic and OpenAI unveiled new, more affordable artificial intelligence models on Tuesday, signaling a strategic pivot as both companies contend with intense rivalry from budget-friendly, open-weight competitors.
OpenAI expanded its GPT-6 lineup with the introduction of GPT-6 Sol and GPT-6 Luna. The company reported that API pricing for these new tiers has been cut by half compared to the promotional rates for GPT-5.6. According to the firm, GPT-6 Sol, positioned below the Astra tier, is designed for complex tasks such as software coding, while GPT-6 Luna targets high-volume operations like document summarization and data extraction.
Simultaneously, Anthropic introduced Claude Opus 5.5, the newest addition to its Claude 5.5 model family. Dianne Penn, Anthropic’s head of product management, research, and labs, explained that the new model is optimized for token efficiency.
“One of the things we’re continuing to innovate on is how to make that thinking, how to make the answering more efficient, so it uses less tokens depending on your effort setting,” Penn told CNBC.
These announcements mark the first product rollouts since Anthropic CEO Dario Amodei advocated for an industrywide slowdown in advanced AI development, a move driven by escalating concerns over model safety. The debate intensified on September 8 when former Anthropic researcher Jacob Coxon announced his departure via X, warning that the labs were “gambling with our lives.” The call for caution was subsequently supported by OpenAI CEO Sam Altman and Tesla and SpaceX CEO Elon Musk.
The push toward cost-effective offerings reflects growing pressure on major AI providers from open-weight models produced by firms such as China’s Alibaba, Moonshot AI, and DeepSeek. Customers are increasingly seeking ways to rein in AI expenditures as they adopt the technology at scale.
Half the price? That sounds like a good deal for my coding projects, but I wonder if quality dropped.