Anthropic, OpenAI cut AI prices with Claude Opus 5.5, GPT-6 Sol and Luna

Both releases came within hours of each other, a sign of the labs' heated rivalry, and they arrive just weeks after Anthropic CEO Dario Amodei published a blog post calling for a coordinated attempt to pace the frontier, which OpenAI CEO Sam Altman supported on X. Anthropic acknowledged Opus 5.5 is its first release since that call, and said the model underwent its most rigorous alignment testing to date, with the same tier of cybersecurity, biology, and frontier-AI-development safeguards previously reserved for Fable 5.1.

The two companies are already deep into a price war playing out on two fronts, according to Ramp lead economist Ara Kharazian: cheaper models businesses are shifting toward, and outright price cuts on the most expensive ones. That dynamic is how technology markets have traditionally worked, he said, though bullish investors may not be factoring it in. Neither Opus 5.5 nor the GPT-6 offshoots are frontier models, but the real fight may be for everyday business customers rather than the most powerful models.

Many businesses are tightening AI budgets and scrutinizing ROI, said Randall Hunt, CTO at the AI consulting company Caylent, noting CFOs have seen sticker shock without the promised gains. How much cheaper models actually win business may hinge on whether enterprises shift everyday workloads to them rather than frontier flagships, and on whether the labs can keep margins intact as prices fall.