On the same day, OpenAI and Anthropic launched "affordable models," igniting a "price war."

The simultaneous launch of lower-priced versions of the flagship AI models by two top U.S. AI labs on the same day signifies that competition in the AI industry has shifted from a simple arms race of capabilities to a two-pronged battle of price and performance.

On Tuesday, Anthropic released its new generation model, Claude Opus 5.5, with a cost reduction of approximately 40% compared to its predecessor. On the same day, OpenAI launched two lower-priced versions of ChatGPT-6, Sol and Luna, priced 50% lower than the ChatGPT-5.6 series. The near-simultaneous releases by both companies, with their highly overlapping timing, clearly illustrate the intense competition in the current AI market.

Behind this price-cutting offensive is the continued pressure from China's "open weight" model—a model that allows users to customize and run on local hardware. Meanwhile, Anthropic is actively preparing for its planned IPO this fall, which is expected to be the largest ever; OpenAI, after sluggish growth at the beginning of the year, is hoping to reignite its growth momentum through price cuts.

Pricing Details: Price Comparison between Anthropic and OpenAI

In terms of pricing, Anthropic's Opus 5.5 costs $4 per million input tokens and $20 per million output tokens , while its current top-performing model, Fable 5.1, is priced at $10 and $50 respectively—on par with OpenAI's flagship model, Astra. Anthropic states that Opus 5.5 performs comparably to Fable 5.1 on most tasks.

On the OpenAI side, Sol is priced at $2 per million input tokens and $10 per million output tokens; the lightweight version, Luna, is priced as low as $0.10 per million input tokens and $0.50 per million output tokens. OpenAI first released its most powerful model, Astra, earlier this month, and Sol and Luna followed suit as lower-priced supplementary versions of the series.

According to reports, after a period of all-out arms race focused on extreme capabilities, the two major U.S. AI labs are increasingly shifting their competitive focus to a combination of price and performance in order to win over cost-sensitive enterprise clients.

"Open weights" models from China pose an increasingly significant competitive threat. According to cloud application company Vercel, these user-customizable, locally run models are taking market share from US labs. In response to this pressure, OpenAI has lowered model prices several times in recent months and claims its Astra model is currently the most powerful model on the market.

OpenAI previously stated that it is unlikely to go public before 2027. After experiencing a slowdown in growth at the beginning of the year, price reductions have become one of its core strategies to revitalize growth.

For Anthropic, the strategic significance of this launch goes beyond price competition. The company is actively expanding its customer base, paving the way for its planned IPO this fall—expected to be one of the largest tech IPOs in history. Investors widely speculate that the release of the new model will be a catalyst for the company to formally file its prospectus.

The core question that the market is closely watching is whether Anthropic can maintain its high growth rate in the coming years, given the increasing competition and more price-sensitive customers.

Safety Commitment: Price Reduction While Emphasizing Risk Management

It is worth noting that this release also occurred against the backdrop of the escalating debate surrounding AI security.

Opus 5.5 is the first new model launched by Anthropic after CEO Dario Amodei publicly called on AI labs to "keep pace with the forefront of development" and to deeply integrate third-party testing organizations. Previously, researchers warned that an unrestrained race for self-improving superintelligence could pose a threat to humanity.

Anthropic stated that the new model underwent "extensive alignment testing, external agency evaluation before release, and appropriate safeguards in high-risk areas such as cybersecurity and biology."

On Tuesday, OpenAI also announced priorities for further collaboration with independent assessment bodies, covering areas such as model training, security, risk preparedness, and incident investigation. This comes just a week after Anthropic announced a partnership with Accenture to conduct security testing.

Risk Warning and DisclaimerInvesting involves risk; please exercise caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Any investment decisions made based on this information are at your own risk.