AI chat with your own API key
A private ChatGPT-style chat for DeepSeek, OpenAI, Claude, Qwen, Kimi and other providers, on a key you bring.
Your key stays in your browser.It goes straight to the provider, never to our servers, and we neither log nor keep it. Use a dedicated key with a spending cap, and delete it in the provider's dashboard when you are done.
Result
- The result will appear here.
Input
A plain chat window for any model your key can reach: DeepSeek, OpenAI, Anthropic Claude, Google Gemini, Qwen, Kimi, GLM, Doubao, OpenRouter, a local Ollama, or any OpenAI-compatible endpoint. It is for people who already pay for API access and want to use it without a monthly subscription on top, or who want to try a model their usual app does not offer. The conversation stays on the page, so follow-up questions carry the earlier turns as context, and replies stream in as the model writes them. The idea is the one ChatGPTNextWeb/NextChat (MIT, 88k+ stars) made popular — a front end for your own key — here without deploying anything.
How it works
- Your browser sends each message straight to the provider you pick, with your key in the request header; our server is not on that path and never sees the key or the conversation.
- The whole conversation so far is re-sent with every new message, because chat APIs are stateless — that is how the model remembers, and why a long chat costs more tokens per reply.
- The system prompt asks for answers in the language you write in, code in fenced blocks, and a plain admission when the model is unsure rather than an invented citation.
- Nothing is stored: the chat lives in this tab until you press New chat or close it, and the key is kept only if you tick the remember box, in this tab's sessionStorage.
Where your data goes
With your own key, what you type and the key go from your browser straight to the provider you choose; neither passes through Hysen Labs. In hosted mode, your text goes through our server to our provider (DeepSeek) and is billed in credits; we keep the token counts and cost of each run for billing, never the text itself or the answer. How the provider handles the text is governed by its own privacy policy.
This tool handles keys and credentials, so nothing about a run is saved, not even to your own history.
About your API key
We do not collect, store or leak your key: it lives only in this page's memory (unless you tick “remember in this tab”) and is gone when you close it. Still, treat any key you have pasted into a web page with care — create a dedicated key with a spending cap for use here, and delete or rotate it in your provider's dashboard when you are done.
What it costs
This tool is free, with no sign-in and no points.
Common questions
- Is this cheaper than a ChatGPT or Claude subscription?
- For light use, usually. You pay the provider per token, and a few dozen questions a day on an inexpensive model costs cents rather than a monthly fee. Heavy use of the top-tier models can cost more than a subscription, so check your provider's price list before settling in.
- Which key should I paste here?
- Create a separate key just for this page, give it a low spending limit if your provider supports one, and revoke it when you are done. The key never reaches us, but a dedicated, capped key means that a leak anywhere else — a browser extension, a shared computer — costs you at most that cap.
- Can the model browse the web or read my files?
- No. It knows what is in this conversation and what it learned in training, nothing more. Paste the text you want it to work with. If you ask about recent events it should tell you it may be out of date rather than guess; for long documents, the summarizer or paper tools give more structured results.
- Why does the request fail with an authentication error?
- A 401 or 403 means the provider rejected the key: it was mistyped, it has been revoked, the account has no balance, or it belongs to a different provider than the one selected. A DeepSeek key will not work against OpenAI, so check that the provider menu matches where the key came from.
The open-source behind it
This tool is a self-contained implementation. ChatGPTNextWeb/NextChat (MIT) does the same job as a library — if you need this behaviour inside your own program, start there rather than calling a web page.
ChatGPTNextWeb/NextChatAlso known as
- ai chat with own api key
- byok ai chat
- chatgpt alternative
- deepseek chat online
- openai api key chat
- nextchat online