TL;DR
Amazon is looking at OpenAI and other alternatives after a renegotiated contract moves Anthropic settlement to a per-token price next year.
Amazon is looking for cheaper alternatives to Anthropic’s Claude models According to The Information, after the renegotiated deal, the company’s AI will move to token-based pricing, which could significantly increase its costs. The new pricing structure doesn’t go into effect until next year, but Amazon is already exploring options including OpenAI. The report highlights a deepening rift between the two companies, once inseparable partners in the AI race.
Amazon’s dependence on Claude runs deep. According to The Information, its coding agent Kiro, workplace assistant Quick and consumer-facing Alexa for shopping are all based on Anthropic’s models. The move to token-based settlement will make this dependency more expensive, especially after Amazon recently abolished the internal leaderboard This encouraged workers to burn as many AI tokens as possible.
The search for cheaper models led Amazon to OpenAI, a company it is already getting closer to. Earlier this year, Amazon committed $50 billion to OpenAI, giving the AI lab access to its cloud infrastructure in exchange for access to its models. The deal follows Amazon’s initial $4 billion investment in Anthropic, which has since grown to a potential $33 billion.
And Anthropic is expanding its connections beyond Amazon. According to The Information, the company has committed to spending $200 billion on Google Cloud and chips over five years. Amazon’s latest $25 billion investment in Anthropic AWS included more than $100 billion in mutual obligations in its spending, but the Google arrangement means Anthropic is no longer dependent on a single cloud provider.
Tensions boiled over the US government last month Ordered Anthropic to shut down Fable 5 and Mythos 5 models After the safety report from Amazon. Andy Jassy reportedly told government officials that Amazon researchers were using Fable 5 to obtain information useful for cyberattacks. As Amazon prepares to launch its own cybersecurity-focused artificial intelligence agent designed to spot vulnerabilities, the timing has raised questions.
The contract dispute, the transition to OpenAI, and the Fable 5 incident together indicate that the Amazon-Anthropic relationship has entered a new and more adversarial phase. Amazon remains one of Anthropic’s biggest investors and cloud customers, but both companies now have reasons to reduce their dependence on each other. For the broader AI industry, the disruption of its most prominent investor-model-provider partnership will redraw the competitive map.
