AI vendors slash token prices to spur enterprise adoption
OpenAI's GPT-6 Sol and Luna models are priced 50% lower per token than their predecessors, while Anthropic's Claude Opus 5.5 costs 20% less than Opus 5. Both companies emphasize efficiency gains from caching and inference improvements. Analysts see this as a price-performance race that could shift enterprise AI deployment strategies.
OpenAI's GPT-6 Sol and Luna carry per-token prices half those of GPT-5.6, while Anthropic's Claude Opus 5.5 is 20% cheaper per token than Opus 5 and costs 40% less to run. Both vendors attribute the reductions to improved caching and inference efficiencies, passing savings to customers. These cuts target production workloads like coding agents and enterprise automation.
Analysts note that lower API costs improve cloud consumption economics but do not erase demand for private or hybrid deployments, given data sovereignty and security needs. They advise CIOs to measure cost per business outcome rather than per token. The pricing pressure may also push hyperscalers toward higher-value services such as orchestration and governance, as raw compute margins shrink.
Lower token prices could democratize access to frontier AI, allowing smaller enterprises to deploy advanced automation previously reserved for large corporations. This may accelerate productivity gains in sectors like software development and customer service, but also raises questions about job displacement and the concentration of AI infrastructure among a few vendors. As costs fall, the emphasis on cost-per-outcome could push providers to prioritize reliability, potentially benefiting end-users through more consistent AI performance.