OpenAI leans on Amazon for enterprise distribution

OpenAI is pitching Amazon Web Services as a new route into big companies, telling staff that Microsoft had constrained some enterprise sales. (cnbc.com) In a memo sent Sunday and viewed by CNBC, chief revenue officer Denise Dresser said many enterprise buyers already use Amazon Bedrock, Amazon Web Services’ marketplace for artificial intelligence models, and called customer demand for the new offering “staggering.” She told CNBC earlier this month that enterprise now accounts for 40% of OpenAI revenue and is on track to match consumer revenue by the end of 2026. (cnbc.com) The Amazon push follows a February 27 partnership announcement: Amazon said it would invest up to $50 billion in OpenAI, while the companies said OpenAI models and a new “Stateful Runtime Environment” would be distributed through Amazon Bedrock. OpenAI also said Amazon Web Services would become the exclusive third-party cloud distribution provider for OpenAI Frontier, its platform for managing teams of artificial intelligence agents. (openai.com) That sales pitch lands after OpenAI and Microsoft rewrote their alliance on October 28, 2025. OpenAI said then that Microsoft would keep Azure application programming interface exclusivity until artificial general intelligence, while some non-application programming interface products could run on other cloud providers. (openai.com) The practical split is becoming clearer in OpenAI’s pricing pages. OpenAI’s help center says that, as of April 2, 2026, ChatGPT Business and ChatGPT Enterprise now have two seat types: a standard ChatGPT seat and a Codex-only seat, both tied to flexible pricing. (help.openai.com) OpenAI has also published a more explicit ChatGPT rate card for business customers. The current table lists GPT-5.4 Thinking at 10 credits per message, GPT-5.4 Pro at 50 credits, Deep Research at 50 credits per task, images at 5 credits per generation, and voice at 5 credits per minute. (help.openai.com) For coding, OpenAI says Codex moved on April 2 from per-message billing to token-based pricing for new and existing ChatGPT Business customers and new Enterprise customers. The current rate card lists GPT-5.4 at 62.5 credits per million input tokens and 375 credits per million output tokens, and says average Codex spending runs about $100 to $200 per developer per month. (help.openai.com) OpenAI’s public business pricing page still keeps the base ChatGPT Business subscription simple at $20 per user per month on annual billing, while Enterprise remains a sales-led product with custom terms. The new rate cards add a second meter on top of that seat price for higher-end models, research tasks, images, voice, and coding workloads. (openai.com; help.openai.com; help.openai.com) The result is a two-track enterprise strategy: Amazon for broader distribution into companies already buying through Amazon Web Services, and published credit tables for finance teams that want clearer unit economics before they expand usage. Microsoft is still central to OpenAI’s stack, but OpenAI is now spelling out more of the sales and pricing machinery around that dependence. (cnbc.com; openai.com; openai.com)