RelayVano API Relay: Claude Setup and 16% Referral Earnings
What the RelayVano referral programme pays
RelayVano is an independent relay made for developers worldwide, and our referral programme turns your own usage into a small, predictable income stream. When someone signs up through your referral link and later adds credit to their account, you earn 16% of every top-up they make. One level further out, the rate is 4%: if a developer you referred brings in a colleague, you still receive a share of that colleague's top-ups without any extra work on your side.
The important detail is that the reward is tied to top-ups, not to sign-ups. A referral who never adds credit earns you nothing, but a referral who keeps paying in becomes a recurring contributor to your balance. That is why the programme works well for developers who already recommend their tools to other engineers, whether that happens in a chat room, a team channel or a code review.
You can withdraw your referral earnings once they reach $5.56, or move them straight into your RelayVano balance and spend them on requests immediately. There is no monthly fee attached to anything we sell, and no subscription to maintain. We simply hold a prepaid balance, you pay per token, and the balance never expires.
New accounts also start with $0.40 of credit, which is enough to send a few real requests and see exactly how the relay behaves before you commit to a larger top-up.
Ten minutes from sign-up to your first request
A quick start on RelayVano is deliberately short. First, create your account and choose your language. The whole site, dashboard and documentation come in six languages: 中文, English, Español, हिन्दी, Português, Français. Prices are shown in yuan on the Chinese pages and in approximate dollars on the others, so a developer reading in any of these languages sees the same service with familiar numbers.
Next, open the dashboard and create one API key. That single key is the credential you use across the platform, because one API key reaches every model through https://api.relayvano.com. The relay exposes an OpenAI-compatible /v1 path, an Anthropic /v1/messages path and the native Gemini /v1beta path, so you can point your preferred client at the base address and keep working in the interface you already know.
Then add credit through Stripe checkout. Cards work everywhere, and WeChat Pay is available where your account supports it. Unspent top-ups can be refunded within 9 days, which keeps the decision to try RelayVano low-risk.
Finally, paste the base address into your client, send a test request and open the usage page. Every request appears there with its charge, itemised so you can see how prompt tokens, reply tokens, cache stores and cache hits were priced. If a caller disconnects mid-stream, the request is still finished and billed on the tokens really used, so nothing disappears from the ledger.
Pairing Claude, GPT and Gemini with the right client
Here is where a ten-minute setup pays off, because RelayVano follows model protocols precisely rather than pretending every model speaks every dialect. Claude models work only in the official Claude Code and Claude Desktop clients, version 2.1.285 or newer. That pairing is exact and without exceptions.
Codex, Cursor, Cherry Studio, Cline, OpenCode, OpenAI SDKs and cURL pair only with GPT, DeepSeek and the other non-Claude, non-Gemini models. Gemini models use the native Gemini protocol only, through Gemini CLI; they are not OpenAI-compatible, so pointing an OpenAI-style client at Gemini will not work as you might expect.
Getting this right takes seconds when you check it against the model list before you start. The practical sequence is simple: decide which model family you want, open the matching client, confirm the client version for Claude is 2.1.285 or newer, then set the base address and your key. RelayVano does the translation work on its side, but the client choice stays yours.
Because the same key and the same base address serve every model, switching families later is not a migration project. You change the model name, and you change the client if the protocol requires it.
Pay-as-you-go fares and what a busy day costs
RelayVano charges per token against a prepaid balance, with no monthly fee and no expiry on what you have paid in. The fares below are approximate US dollars per million tokens, using 1 USD ≈ ¥7.2, and the charge itself is made in yuan.
For a feel for the numbers, Claude Sonnet 5.5 is listed at $1.25 prompt, $6.25 reply and $0.0889 cache hit. GPT-6.1 Sol is $0.1958 prompt, $0.9792 reply and $0.0389 cache hit, while Gemini 3.8 Flash is $1.20 prompt, $5.40 reply and $0.1194 cache hit. Models such as MiniMax M3 ($0.1472 prompt, $0.5889 reply, $0.0292 cache hit) sit at the lighter end of the range.
Here is an example rather than a promise: suppose you send 200 requests a day, each with 1,000 prompt tokens and 500 reply tokens, using Claude Sonnet 5.5. The prompt side works out at $0.25 for the day, the reply side at $0.625, and the total at $0.875. Over 30 days that example becomes $26.25 drawn from your prepaid balance. Change the model and the same workload shifts accordingly, because every token category is priced separately and shown separately.
Cache hits are worth watching. On Claude Sonnet 5.5, a cache hit is $0.0889 against a $1.25 prompt rate, so reusing context can meaningfully reduce a repeated workload's cost. Your usage page lets you confirm whether that is happening in practice.
Growing the referral link into withdrawable earnings
Once your own setup is comfortable, share your referral link with developers you know. Every top-up they make through it adds 16% to your earnings, and their own referrals add 4% at the second level. The rate applies to top-ups, so the value compounds with how actively your referrals use the service, not with how many people click.
The threshold is $5.56, and you have two ways to use the balance once you cross it: withdraw it, or move it into your RelayVano balance and let it pay for your own requests. Moving it in is often the simpler choice for a developer who is already spending on tokens each day, because it turns the referral programme into a discount on your own work.
It is also worth keeping your own top-ups modest while you measure. Unspent top-ups can be refunded within 9 days, so a smaller first purchase gives you time to check the usage page, compare the fares for the models you actually call, and see how the referral earnings build before you scale up.
Six-language support and a straightforward place to ask
If something needs checking, logged-in members write to us under Messages in the dashboard. That keeps support inside the same place as your key, your balance and your usage history, so we can look at the request you are asking about rather than asking you to reconstruct it.
RelayVano was built for developers worldwide, and the six-language site, dashboard and docs reflect that. Prices appear in yuan on the Chinese pages and in approximate dollars on the others, while the underlying billing stays in yuan. One API key, one base address, per-token billing and a referral programme that pays 16% on the top-ups of the people you invite: that is the whole offer, arranged so you can start in ten minutes and grow from there.
Create your RelayVano account, claim the $0.40 starter credit, send your first request and pick up your referral link. Invite the developers you work with, and let their top-ups start paying you back.