Gemini CLI on RelayVano: native API relay setup and pricing

Coding-agent routes EN2026-11-01

Why Gemini CLI and RelayVano use the native Gemini route

Gemini CLI was built to speak Gemini's own protocol, so that is the route we recommend when you connect it to RelayVano. Gemini models use the native Gemini protocol only, which means Gemini CLI is the client that fits them. They are not OpenAI-compatible, and pretending otherwise tends to produce confusing errors or a request that never reaches the model.

RelayVano keeps that distinction clear. One API key reaches every model through https://api.relayvano.com, but the client you choose still has to match the model family. For Gemini models, you stay with Gemini CLI and its native route. Other clients in our lineup pair with GPT, DeepSeek and the other non-Claude, non-Gemini models, while Claude models work only in the official Claude Code and Claude Desktop clients, version 2.1.285 or newer.

That sounds restrictive, but in practice it removes guesswork. You are not translating a Gemini request into an OpenAI-style payload and hoping the result behaves as expected. Gemini CLI asks for Gemini, RelayVano forwards the request on the native Gemini route, and the response comes back in the format the client already understands.

We also built RelayVano for developers worldwide. The whole site, dashboard and docs come in six languages: 中文, English, Español, हिन्दी, Português, Français. Prices show in yuan on the Chinese pages and in approximate dollars on the others, so the same account can be used comfortably from different parts of the world without hunting for a separate guide in another language.

Setting up Gemini CLI with one RelayVano key

The setup is short because RelayVano is designed around a single API address. First, create your RelayVano account and add credit with Stripe checkout. Cards work everywhere, plus WeChat Pay where your account supports it. Unspent top-ups can be refunded within 9 days, which gives you room to try the service before committing more of your budget.

Next, copy the API key from your dashboard. Open your Gemini CLI settings, point the client at our base address, https://api.relayvano.com, and keep Gemini CLI's native Gemini protocol configuration. Insert the RelayVano key in the place where Gemini CLI expects its credential. There is no separate Gemini account to connect and no second key to manage for this route.

Then select one of the Gemini models offered on RelayVano and send a small test prompt. A short request is a sensible way to confirm that the address, key and model name are all in the right places before you start a larger task. Once the reply arrives, open the usage page in the RelayVano dashboard. Every request appears there with its charge, so you can check the first entry against the fare you expected.

If you use the same RelayVano key from another supported client later, remember the pairing rule again. The key is shared, but the protocol is not. Gemini CLI is the right home for Gemini models. Codex, Cursor, Cherry Studio, Cline, OpenCode, OpenAI SDKs and cURL pair only with GPT, DeepSeek and the other non-Claude, non-Gemini models.

Worked example: 200 Gemini CLI requests a day

Here is a worked example, clearly labelled as an example. Assume you send 200 requests a day through Gemini CLI, and each request contains 1,000 prompt tokens and produces 500 reply tokens. We will price it with Gemini 3.1 Pro, whose approximate RelayVano fares are $1.58 per million prompt tokens and $9.13 per million reply tokens. The charge itself is made in yuan.

Prompt tokens per request: 1,000 divided by 1,000,000, multiplied by $1.58, which equals $0.00158.
Reply tokens per request: 500 divided by 1,000,000, multiplied by $9.13, which equals $0.004565.
Cost per request: $0.00158 plus $0.004565, which equals $0.006145.
Cost for 200 requests in one day: 200 multiplied by $0.006145, which equals $1.229.
Cost for 30 days: 30 multiplied by $1.229, which equals $36.87.

This is an example for illustration, not a promise about your own bill. Your real cost depends on the model you select and the tokens your prompts and replies actually use. The useful part of the calculation is the method: take each request, split it into prompt tokens and reply tokens, price the two parts separately, then multiply by the number of requests.

Because RelayVano bills per token with a prepaid balance, you do not need to estimate a subscription-sized commitment before you begin. There is no monthly fee, and your balance never expires. That means the $36.87 figure in this example can be funded in a top-up and spent at the pace your work actually requires.

A lighter Gemini 3.8 Flash example for repeated short tasks

Here is a second worked example, also clearly an example. Suppose the same 200 requests a day each use 1,000 prompt tokens and 500 reply tokens, but you run them on Gemini 3.8 Flash. Its approximate RelayVano fares are $1.20 per million prompt tokens and $5.40 per million reply tokens.

Prompt tokens per request: 1,000 divided by 1,000,000, multiplied by $1.20, which equals $0.0012.
Reply tokens per request: 500 divided by 1,000,000, multiplied by $5.40, which equals $0.0027.
Cost per request: $0.0012 plus $0.0027, which equals $0.0039.
Cost for 200 requests in one day: 200 multiplied by $0.0039, which equals $0.78.
Cost for 30 days: 30 multiplied by $0.78, which equals $23.40.

Again, this is an example rather than a quotation for your project. It shows how a different Gemini model changes the arithmetic without changing your workflow. You still use Gemini CLI, the same base address and the same RelayVano key; only the selected model and the resulting per-token fares move.

Longer conversations can also involve cached content. RelayVano prices prompt tokens, reply tokens, cache stores and cache hits separately, so a request that reuses cached material is not simply charged as if every token were fresh. We do not hide those components behind a flat monthly allowance. The usage page breaks the charge down, which makes it easier to see whether a prompt grew, a reply grew or a cache hit reduced the cost.

One more billing detail matters when you run long generations. If a caller disconnects mid-stream, the request is still finished and billed on the tokens really used. You are not charged for a full reply you never received, and you are not asked to pay for work the model never performed. The tokens actually consumed are the ones counted.

What you pay for, and how the balance works

RelayVano sells pay-as-you-go balance and nothing else. You top up, we deduct per-token charges as requests arrive, and there is no monthly fee. The balance never expires, so a small leftover amount is still there when you return to a project weeks later. New accounts start with $0.40 of credit, which is enough to make a first Gemini CLI request and see how the usage page records it.

Top-ups run through Stripe checkout. Cards are accepted everywhere, and WeChat Pay is available where your account supports it. If you change your mind about unspent top-ups, you can request a refund within 9 days. We would rather you start with a small balance, check the numbers, and then decide how much to add than ask you to buy a large commitment up front.

If you enjoy the setup and want to bring colleagues along, our referral programme pays 16% of every top-up by people who sign up through your link, and 4% one level further. Earnings can be withdrawn from $5.56 or moved into your balance, where they are immediately usable for requests.

Questions are easy to ask. Once you are logged in, write to us under Messages in the dashboard, and a member of the team can look at your setup, model choice or usage entries with you.

Start your first Gemini CLI request on RelayVano

Gemini CLI on RelayVano comes down to a few decisions: create an account, add a small amount of credit, point Gemini CLI at https://api.relayvano.com with your RelayVano key, keep the native Gemini protocol, and choose a Gemini model. Send one short prompt, then check the usage page to confirm the charge matches your expectation.

From there, the same key and the same prepaid balance fit into a wider workflow. The site, dashboard and docs are available in 中文, English, Español, हिन्दी, Português and Français, and the per-token pricing stays visible rather than hidden inside a subscription. Sign up for RelayVano today and put Gemini CLI to work on your own terms.