Anthropic and OpenAI launch cheaper models amid controversy over AI safety

Anthropic and OpenAI announced cheaper versions of their artificial intelligence models on the same day, September 22, in a move that responds to mounting competitive pressure from rivals offering open-weight models at far lower prices. These are the first releases from either lab since an intense debate erupted over the existential risks of advanced AI.

OpenAI expanded its GPT-6 family with two new tiers: GPT-6 Sol and GPT-6 Luna, cutting API prices by 50% compared with the promotional pricing for GPT-5.6. Sol sits one rung below Astra and is designed for complex workloads such as coding, while Luna is aimed at "high-volume tasks" like information extraction or document summarization.

Anthropic, for its part, introduced Claude Opus 5.5, the newest addition to its Claude 5.5 family. The company describes it as a more token-efficient version, with running costs roughly 40% lower than Opus 5. Dianne Penn, head of product management, research and labs at Anthropic, told CNBC that one of the company's areas of innovation is making "reasoning" and response generation more efficient, consuming fewer tokens depending on the configured effort level.

The backdrop to these launches is significant: they come after Anthropic CEO Dario Amodei's call for an industry-wide slowdown in advanced AI development, amid a debate over model safety that peaked in the preceding weeks. That debate was triggered after a former Anthropic researcher, Jacob Coxon, posted on X on September 8 that he was leaving his job, warning that both labs were "gambling with our lives." OpenAI CEO Sam Altman and Tesla and SpaceX CEO Elon Musk also joined Amodei's call.

Beyond the safety debate, the article notes that these launches also answer a very concrete commercial need: satisfying customers who want more cost-effective models and are trying to rein in their AI spending. Both Anthropic and OpenAI face pressure from competitors offering cheaper open-weight models, among them the Chinese firms Alibaba, Moonshot AI and DeepSeek.

Taken together, the news reflects an underlying tension in the AI industry: while the sector's leading voices urge caution and warn of long-term safety risks, commercial competition — especially price pressure from open models, many of them Chinese — is pushing the leading labs to keep shipping cheaper, more accessible products.