Frequently asked questions

What is 6bee.ai?

6bee.ai is a prepaid API for open-weight, uncensored large language models. You point Claude Code, Codex or any OpenAI-compatible tool at one base URL, https://api.6bee.ai, and pay per token from a prepaid balance. There is no subscription.

It is aimed mainly at coding tools and AI agents, so the limits are sized for long sessions.

What does “uncensored” mean on 6bee.ai?

Every model listed is an open-weight model that has been tuned to remove refusals. In practice it refuses far fewer requests than a mainstream assistant, without losing capability.

It does not mean anything goes. The Acceptable Use rules apply regardless of what a model is willing to generate. See what uncensored means here.

Which models can I use?

All models currently listed are open-weight models. The current list, with prices and context limits, is in the console and at GET https://api.6bee.ai/v1/models. Closed models such as Claude or GPT are not listed.

Is 6bee.ai compatible with the OpenAI API?

Yes. One host speaks three formats: OpenAI Chat Completions (POST /v1/chat/completions), OpenAI Responses (POST /v1/responses) and Anthropic Messages (POST /v1/messages). All three stream over server-sent events and share one key, one balance and one set of limits.

“Compatible” means the request format is broadly the same. Not every parameter of other APIs is supported. Details are on the API formats page.

Is 6bee.ai affiliated with Anthropic, OpenAI or the model makers?

No. Claude Code and Codex are tools you can point at 6bee.ai because it speaks the same API formats. 6bee.ai is not affiliated with or endorsed by them or by any upstream provider. Model names and trademarks belong to their owners.

Is the code open source?

The gateway and billing code that runs the console and API is public under AGPL-3.0, so anyone can check how tokens are counted and billed. The exact version in production, the patch and the build steps are on the source page.

Is there an uptime guarantee?

Not during the beta. The service is provided on a best-effort basis with no SLA. Models, prices and limits can change, including when an upstream provider changes its service.

Do you store my conversations?

No. Prompts, context and completions are not saved, documents and codebases are not scanned, and request content is not written to logs. They pass through our systems only to be forwarded to the model and returned to you.

What we do keep is what billing needs: your account, your top-ups and per-request usage metadata (model, token counts, cost, time, IP address, status). See what is and is not stored.

Are my prompts used to train models?

Not by us. We do not store your content, so we have nothing to train on.

Who else can see my prompts?

To generate a response, your prompt is sent to whoever serves the model you chose: a third-party inference service or a server we run ourselves. A third-party service handles it under its own terms. Some providers also receive an anonymous code for your account, used for abuse controls and cache isolation. It contains no email address or username.

To learn who serves a particular model, email support@6bee.ai.

Do I need to verify my identity?

No. Sign-up takes an email address and a CAPTCHA check. We do not ask for your real name, a phone number or identity documents, and we run no KYC checks. You can also pay in USDT or USDC.

Where are your servers?

The application server and database are in Germany (EU). Traffic passes through Cloudflare’s network. Upstream model providers may be located in other countries.

Do you use tracking or analytics?

We run our own first-party visit statistics to learn which channel visitors come from. It stores hashed anonymous IDs, not IP addresses. There are no advertising trackers and no third-party analytics platform. The Privacy Policy lists the two cookies involved.

How do I pay?

You top up a prepaid balance in the console. You can pay by bank card (processed by Stripe; we never see the card number) or in USDT or USDC. Prices are in US dollars.

How does paying with USDT or USDC work?

USDT and USDC are charged 1:1 with the USD price. The order page shows the network, the address and the exact amount. Send exactly that amount on that network. On-chain transfers cannot be reversed, so a wrong network, address or amount may be unrecoverable. More on the crypto payment page.

Is there a subscription or a monthly fee?

No. It is prepaid and billed per token. There is no subscription, no credit line and no automatic recharge.

How is usage billed?

Your balance is charged per token at each model’s price. Each request first reserves its input plus max_tokens of output. When it finishes you pay for what was actually used, and the rest goes back to your balance.

Can I get a refund?

Credits are a digital good and are delivered when the top-up is credited, so consumed credits and unused balance are not refunded. The exceptions are a duplicate charge or a payment that was never credited, us ending the service or closing your account without cause, and refunds the law requires. See the Refund Policy.

How much does it cost?

Each model has its own per-token price, shown in the console. You only pay for tokens used.

How do I use 6bee.ai with Claude Code?

Set ANTHROPIC_BASE_URL to https://api.6bee.ai, ANTHROPIC_AUTH_TOKEN to your key and ANTHROPIC_MODEL to a model ID, then run claude. The Claude Code guide has the full setup.

How do I use 6bee.ai with Codex?

Add a model provider with base URL https://api.6bee.ai/v1 and wire_api = "responses" to ~/.codex/config.toml. The Codex guide has the file to copy.

Does it work with Cherry Studio, Chatbox, Cline and other apps?

Yes. Add 6bee as an OpenAI-compatible provider with base URL https://api.6bee.ai/v1, your key and a model ID. There are step-by-step guides for Cherry Studio, Chatbox, Cline and Roo Code and OpenCode.

What are the rate limits?

During the beta, per account across all of its keys: 60 requests per minute, 2 requests at a time, and 1,000,000 uncached input tokens per rolling minute (cache hits do not count). A request body can be up to 8 MiB.

How large can the context and the output be?

Input plus output up to the model’s window: 128K to 1M tokens, depending on the model. For output: max_tokens up to the model’s output cap: 16,384 to 131,072, depending on the model. Larger values are lowered to the cap, not rejected. If omitted: 8,192 to 32,768.

What happens when I go over a limit?

Over a limit you get HTTP 429 with Retry-After, or 400 when a request can’t fit the context. A lowered output cap is reported in the X-6bee-Output-Cap response header; on /v1/messages a thinking budget is lowered with it.

What am I not allowed to do?

The Acceptable Use rules in the Terms of Service apply whatever a model will generate. Among other things they prohibit illegal activity, any sexual content involving minors, non-consensual intimate imagery of real people, malware and unauthorized attacks, fraud and phishing, harassment and threats, help with weapons capable of mass casualties, and reselling access or sharing keys without permission.

Can I use 6bee.ai from mainland China?

No. The service is not offered to users in mainland China, or in countries or regions under comprehensive sanctions.

How do I get help?

Message @newbeeai on Telegram or email support@6bee.ai. To report abuse, email abuse@6bee.ai.