OpenAI and Anthropic publish enterprise price sheets

OpenAI and Anthropic have both put more of their business pricing in public view, giving corporate buyers clearer numbers to compare before they call sales. (help.openai.com) (platform.claude.com) OpenAI’s new ChatGPT rate card for Business and Enterprise or Edu says GPT-5.4 Thinking costs 10 credits per message, GPT-5.4 Pro costs 50 credits, agent mode costs 30 credits per message, deep research costs 50 credits per task, images cost 5 credits per generation, and voice costs 5 credits per minute. The Help Center article says it was updated two days ago. (help.openai.com) The same OpenAI article says GPT-4o, GPT-4.1, GPT-4.1 mini, OpenAI o4-mini, and GPT-5 Instant and Thinking were retired from ChatGPT on February 13, 2026, with GPT-4o in Custom GPTs fully retired after April 3, 2026. OpenAI’s pricing page lists ChatGPT Business at $25 per user per month on annual billing and says Enterprise pricing is available through sales. (help.openai.com) (chatgpt.com) OpenAI also said Codex, its coding agent, is included with ChatGPT Plus, Pro, Business, and Enterprise or Edu plans, and “for a limited time” it is included with Free and Go. The company says Codex can edit files, run commands, execute tests, and review pull requests in GitHub. (help.openai.com) Anthropic’s pricing page lays out token charges model by model, which is how developers pay for text sent into and out of a model, like a metered cloud bill. Claude Sonnet 4.6 is listed at $3 per million input tokens and $15 per million output tokens, Claude Haiku 4.5 at $1 and $5, and Claude Opus 4.6 at $5 and $25. (platform.claude.com) Anthropic also published separate prices for prompt caching, a feature that stores reused parts of a prompt so later requests can run cheaper and faster. Its page lists cache hits for Claude Sonnet 4.6 at $0.30 per million tokens and says Google Vertex AI regional and multi-region endpoints carry a 10 percent premium over global endpoints for Claude 4.5 models and later. (platform.claude.com) The practical change is that OpenAI is showing more of the unit economics inside ChatGPT seats, while Anthropic is showing more of the unit economics inside its application programming interface. That gives procurement teams and software buyers a cleaner side-by-side than the older pattern of custom quotes and scattered documentation. (help.openai.com) (platform.claude.com) The two price sheets are still not identical products. OpenAI’s published numbers focus on workplace subscriptions, credits, and bundled tools inside ChatGPT, while Anthropic’s page focuses on direct developer access priced per token across Claude models and features. (help.openai.com) (platform.claude.com) That leaves sales negotiations in place for the biggest accounts, but with fewer blank spaces at the start of the conversation. The new documents give buyers a public baseline for what a message, a task, a minute of voice, or a million tokens costs before discounts and volume terms enter the picture. (chatgpt.com) (help.openai.com) (platform.claude.com)