Question 1

What is an OpenAI-compatible API?

Accepted Answer

An OpenAI-compatible API is an HTTP endpoint that accepts the same request bodies and returns the same response shapes as OpenAI’s own API, so an official OpenAI SDK works against it after you change one setting — the base URL. It is a convention, not a standard: there is no certification, and providers implement different subsets of it.

Question 2

How do I switch from the OpenAI API to another provider without rewriting my code?

Accepted Answer

Change the base URL and the model id. If you use the official OpenAI SDK, those are the only two things in your code that have to change — everything else about chat completions, streaming and tool calling stays as it is. The work is in the things around it: endpoints your new provider does not serve, and evals to confirm the new model is good enough for your task.

Question 3

Is there a free LLM API that does not need a credit card?

Accepted Answer

Yes. Our free key needs an email or a Google sign-in, no card and no identity check, and it comes with a weekly allowance worth roughly 3.9M tokens on DeepSeek-V4-Flash. Several vendors run free tiers too; the practical differences are the size of the allowance, whether the free tier is rate-limited, and what happens when you run out.

Question 4

How many tokens does one dollar buy?

Accepted Answer

At a 4:1 input-to-output ratio — roughly what agent and chat traffic runs at — one dollar buys about 3.5M tokens on MiniMax-M2.7, 1.7M on Kimi K2.6, and about 0.3M on Claude Sonnet 5. The spread between an open-weight model and a frontier one is roughly an order of magnitude, which is why the same budget behaves so differently depending on what you point it at.

Question 5

Why is the API returning 429, and what should I do?

Accepted Answer

A 429 means you have spent a rolling cap, not that you are banned. Our 429 comes with a Retry-After header saying how many seconds until the cap frees up — respect it, and the request succeeds. If you are seeing 402 instead, that is a different thing: no allowance and no balance.

Question 6

Which model should I use for a coding agent?

Accepted Answer

Start with the cheapest model that passes your evals, and escalate individual tasks rather than the whole workload. Coding agents spend most of their tokens on reading files, tool results and retries — work where a frontier model buys you very little — so the model choice that matters is per step, not per project.

Straight answers