# Munito AI partners with Opper AI

URL: https://munito.ai/blog/live-on-opper

2026-11-16

Our open models answer through Opper AI, the European AI gateway. Five models, three tiers each, and half price on all of them there until 12 December 2026.

Munito AI is happy to partner with [Opper AI, the European AI gateway](https://opper.ai): one key, one API, more than 700 models, hosted in the EU. Five of our models answer there, and reaching them is a model string.

## Half price on Opper AI, until 12 December 2026

For the first month of the partnership, every Munito model on Opper AI costs half its posted price, whichever tier you pick and wherever the route runs. The offer belongs to the gateway: it rides on the Opper AI key, not on an account with us.

```ts
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.OPPER_API_KEY,
  baseURL: "https://api.opper.ai/v3/compat",
});

const completion = await client.chat.completions.create({
  model: "munito/glm-5.3-flash",
  messages: [{ role: "user", content: "Hello" }],
});
```

## Five models, three tiers

DeepSeek V4 Flash, DeepSeek V4.1 Flash, GLM-5.3-Flash, Qwen3.8 27B and Aleph Alpha's Kolibri 1, each with its full context window. The tier is part of the model string:

| the model string | what it is for |
| --- | --- |
| `munito:flex/glm-5.3-flash` | batch work and evaluations that can wait, at half the standard price |
| `munito/glm-5.3-flash` | the everyday call |
| `munito:priority/glm-5.3-flash` | interactive traffic, served first |

## The EU route is the default

`munito/glm-5.3-flash` runs in EU member states. `munito:us/qwen3.8-27b` runs in the United States, and `munito:global/glm-5.3-flash` runs wherever the capacity is cheapest. On every host we keep no prompts and no outputs, and we train on nothing you send.

Take an Opper AI key and send a request, or come straight to us: [open-model inference](https://munito.ai/inference) and [what it costs](https://docs.munito.ai/pricing).

## About Opper AI

Opper AI is the European AI gateway for agents, and it is based in Stockholm. One key reaches more than 700 models through an API hosted in the EU, pay as you go, under the rules an organization sets: spend caps, model allowlists and zero data retention. It answers the OpenAI SDK, so a project reaches any of those models without a rewrite. Our own routes are on [our provider page at Opper AI](https://opper.ai/provider/munito).
