RelayVano Pay-as-you-go API: MiniMax M3 and Hunyuan 4 Chat Setup

Model atlas EN2026-10-31

MiniMax M3 and Hunyuan 4 on one pay-as-you-go relay

RelayVano is an independent API relay built for developers worldwide. Our whole site, dashboard and documentation are available in six languages: 中文, English, Español, हिन्दी, Português and Français. Prices are shown in yuan on the Chinese pages and in approximate dollars on the others, so a developer reading the docs in Delhi, Madrid, São Paulo, Paris or Beijing sees the same relay with pricing in familiar terms. If you are building a lightweight chat feature — a support widget, a summariser, a small assistant panel — RelayVano gives you a straightforward way to reach MiniMax M3 and Hunyuan 4 without juggling separate accounts for each vendor.

One API key reaches every model through https://api.relayvano.com, using the OpenAI-compatible /v1 path, the Anthropic /v1/messages path and the native Gemini /v1beta path. MiniMax M3 and Hunyuan 4 are non-Claude, non-Gemini models, so they pair with Codex, Cursor, Cherry Studio, Cline, OpenCode, OpenAI SDKs and cURL. The pairing rules are exact and have no exceptions: Claude models work only in the official Claude Code and Claude Desktop clients (version 2.1.285 or newer), while Gemini models use the native Gemini protocol only, in Gemini CLI, and are not OpenAI-compatible. For the two chat models in this guide, the familiar OpenAI-compatible route is the one to use.

A worked example: 200 lightweight chat requests a day

Here is an example, and every sum below is worked from the current RelayVano fares. Assume your product sends 200 requests a day, and assume each request carries 1,000 prompt tokens and receives 500 reply tokens. That gives a daily volume of 200,000 prompt tokens and 100,000 reply tokens. Cache stores and cache hits are priced separately, so we leave them out of this simple example and add them only if your workload actually caches.

MiniMax M3 prompt tokens: 200,000 tokens is 0.2 million tokens. At $0.1472 per million prompt tokens, 0.2 × $0.1472 = $0.02944 per day.
MiniMax M3 reply tokens: 100,000 tokens is 0.1 million tokens. At $0.5889 per million reply tokens, 0.1 × $0.5889 = $0.05889 per day.
MiniMax M3 daily total: $0.02944 + $0.05889 = $0.08833. Over 30 days, $0.08833 × 30 = $2.6499.
Hunyuan 4 prompt tokens: 0.2 million tokens at $0.3514 per million gives 0.2 × $0.3514 = $0.07028 per day.
Hunyuan 4 reply tokens: 0.1 million tokens at $1.12 per million gives 0.1 × $1.12 = $0.112 per day.
Hunyuan 4 daily total: $0.07028 + $0.112 = $0.18228. Over 30 days, $0.18228 × 30 = $5.4684.
Running both models side by side in the example: $0.08833 + $0.18228 = $0.27061 per day, and $0.27061 × 30 = $8.1183 over 30 days.

These figures are approximate US dollars, with the actual charge made in yuan. The point of the example is visibility: in a chat product where replies are short and prompts are compact, the per-request cost is small enough that you can reason about it directly instead of negotiating a contract. If a vendor subscription would cost you more than the traffic you actually generate, pay-as-you-go billing on RelayVano lets you keep spend proportional to use.

Setting up MiniMax M3 and Hunyuan 4 in your own client

Setup takes a few sentences rather than a code listing. First, create your RelayVano account and copy the API key from the dashboard. Next, open the client you already use — Cursor, Cherry Studio, Cline, OpenCode, Codex, an OpenAI SDK or cURL — and point its OpenAI-compatible base address at https://api.relayvano.com/v1. If your client reads an environment variable for the base address, set that variable to the same value and set the API key variable to the key you copied.

Then choose the model. Enter MiniMax M3 when you want a compact, low-cost assistant for short exchanges, and enter Hunyuan 4 when you want a slightly larger reply budget for longer answers. Because both models travel the OpenAI-compatible route, switching between them is a model-name change, not a rewrite of your integration. Send one test message, confirm the reply arrives, and check the usage page to see the charge recorded for that request.

RelayVano does not ask you to learn a second billing system for each model. The same key, the same base address and the same dashboard cover both, and the same prepaid balance covers everything you call. That matters when you are shipping a small feature rather than a platform: fewer moving parts between your code and your monthly bill.

How RelayVano prices tokens and records every request

RelayVano sells pay-as-you-go balance and nothing else. There is no monthly fee, and your prepaid balance never expires, so unused credit is not lost while you are still deciding which model to use for a feature. You are billed per token, with prompt tokens, reply tokens, cache stores and cache hits priced separately, so you can see which side of the exchange is driving cost.

Every request appears on the usage page with its charge attached. That line-by-line record is what makes the worked example above verifiable: you can compare your own prompt and reply totals against the same arithmetic we used, and adjust the model if one side is heavier than expected. If a caller disconnects mid-stream, the request is still finished and billed on the tokens really used, which keeps your cost model honest even when a user closes a tab or a mobile connection drops.

Because the balance is prepaid and per-token, you can start with a small top-up, watch the usage page for a week, and scale only after you have real numbers for your own traffic. For a chat feature, that is usually a more comfortable first step than committing to a fixed spend in advance.

Top-ups, refunds, referrals and support in six languages

Top-ups run through Stripe checkout. Cards work everywhere, and WeChat Pay is available where your account supports it, which covers a good share of the developers we serve. If you change your mind about an unspent top-up, you can request a refund within 9 days. New accounts start with $0.40 of credit, enough to send a handful of real requests and see the usage page in action before you add funds.

RelayVano also rewards developers who bring colleagues along. If someone signs up through your referral link, you earn 16% of every top-up they make, and 4% one level further down. Earnings can be withdrawn from $5.56 or moved straight into your RelayVano balance to cover usage. If you need help, logged-in members can write to us under Messages in the dashboard, and our support team works in the same six languages as the rest of the site.

Where MiniMax M3 and Hunyuan 4 fit your chat feature

Lightweight chat features have a particular shape: short prompts, short replies, many requests, and a cost that scales with traffic rather than with a fixed commitment. MiniMax M3, at $0.1472 per million prompt tokens and $0.5889 per million reply tokens, suits that shape especially well. Hunyuan 4, at $0.3514 per million prompt tokens and $1.12 per million reply tokens, gives you a second option for conversations that need a bit more room, still billed per token through the same key.

RelayVano keeps the surrounding machinery simple: one API key, one base address, one prepaid balance that never expires, per-request charges on a usage page, and documentation in six languages for a worldwide developer audience. If you want to see your own numbers rather than our example, sign up for RelayVano, add a small top-up, point your client at MiniMax M3 or Hunyuan 4, and read the first charge on your usage page. That is the fastest way to know whether these two models fit your chat feature.