Claude API vs OpenAI API: Same 30-Day Clock, Different Ceilings

People compare these two APIs on which model writes better, and no page on either site will settle that. What the pages do settle, in almost identical words, is the part a client cares about: who owns the output, whether the input trains the model, and how long it is kept. Read side by side, the two agreements agree. Then the price ladders and the ceilings diverge, and that is where the choice actually lives.

Where they read the same

Both assign the output to you. OpenAI's services agreement says the customer owns all Output, that OpenAI assigns to the customer all of its right, title and interest in it, and that the customer keeps ownership of Input. Anthropic's commercial terms say the customer owns the Outputs and keeps all rights to Inputs, and that Anthropic assigns its interest in Outputs.

Both say API data is not training data by default. OpenAI's data guide: as of March 1, 2023, data sent to the API is not used to train or improve OpenAI models unless you explicitly opt in. Anthropic's terms: Anthropic may not train models on Customer Content.

Both run a 30-day clock. Anthropic deletes API inputs and outputs within 30 days of receipt or generation, with exceptions for the Files API, a zero data retention agreement, Usage Policy enforcement and the law. OpenAI keeps abuse-monitoring logs, which may contain prompts and responses, for up to 30 days by default, and deletes all Customer Content within thirty days of the agreement terminating. Both offer a zero-retention route, and on both it is an approval or an agreement rather than a checkbox. Neither hands anything back when you leave.

Both can suspend without much warning. Anthropic may suspend all access over a suspected Usage Policy breach and disclaims liability for the loss that causes; OpenAI may limit or suspend access on a breach of the agreement or its policies, without prior notice where necessary.

Where the price ladders diverge

OpenAI starts lower. Its cheapest listed models are gpt-5-nano at $0.05 in and $0.40 out per million tokens and gpt-4o-mini at $0.15 and $0.60. Anthropic's floor is Claude Haiku 4.5 at $1 and $5. In the middle, OpenAI's gpt-5.6-luna is $0.20 and $1.20 and gpt-5-mini $0.25 and $2.00, while Claude Sonnet 5 is $2 and $10, a price the page now calls standard after the scheduled rise to $3 and $15 was cancelled. At the top the two ladders meet: Claude Fable 5.1 is $10 and $50, and so is gpt-6-astra. OpenAI's gpt-5.6-sol sits at $4 and $20 on a promotional price the page says is available at least through November 21, 2026.

So for a job that fits a small model, OpenAI's floor is roughly eight times cheaper on output; for a job that needs the top of either ladder, the price is the same $50. The drafting page shows what those numbers mean per 1,500 words, and the AI job cost calculator runs any size.

Where the ceilings diverge

This is the difference that stops a product, and the two companies build it differently.

OpenAI caps a new account by usage tier. The Free tier, for users in an allowed geography, allows $100 of usage a month. Tier 1 unlocks at $5 paid and also allows $100; Tier 2 at $50 paid allows $500; Tier 3 at $100 paid allows $1,000; Tier 4 at $250 allows $5,000; Tier 5 at $1,000 allows $200,000. Underneath, rate limits are measured in requests and tokens per minute and per day and bite on whichever comes first. The Batch API is the pressure valve: a 50% cost discount, a separate pool of substantially higher rate limits, and completion within 24 hours.

Anthropic caps by spend tier: Start $500 a month, Build $1,000, Scale $200,000, and no cap on Custom. When an organization reaches its cap, API usage pauses until 00:00 UTC on the first day of the next month unless a higher limit is requested; requests return HTTP 429 with the error code enforced_spend_limit_reached and no retry-after header, and the page says retries, including the SDKs' automatic ones, fail until access resumes. Anthropic's pricing page refers to a Batch API discount; this page does not quote a figure for it because the multiplier was not on the page we read.

For a one-person operation the practical difference is small in dollars and large in behaviour. A $5 payment moves OpenAI off the Free tier. An Anthropic Start account that reaches $500 on the 20th is a paused account for ten days. The support bot page walks through what that means for something customers rely on.

What to do today

Decide the question you are actually asking. If it is the client's question, ownership, training, retention, the two agreements give the same answers and you can quote either. If it is the bill, start at the bottom of OpenAI's ladder and the bottom of Anthropic's and compare $0.60 to $5 per million output tokens on your real job. If it is uptime, read the ceiling: $5 paid and a monthly usage limit on one side, a monthly spend cap and a pause until the 1st on the other. Then save your outputs somewhere that is not either API, because both delete in thirty days.