Home›Companies›Munito AI
Munito AI

Open-model inference provider for Europe, private by default.

Munito is an open-model inference provider for Europe, private by default. Its API lets users choose their models and keep inference and storage in the EU. Founded by the former inference lead at Mistral AI, Munito is built on one belief: open and independent intelligence, both models and inference, is key to economic growth and prosperity.
Active Founders
Alexey Eryshev
Alexey Eryshev
Founder/CEO
Founder at Munito AI. I led inference at Mistral AI, where my team built its API and inference products, and I built Le Chat's Deep Research. Before that, I spent five years at Airbnb, where I built Hosting Services from scratch and grew it to a global product.
Company Launches
Munito AI: Open-model inference provider for Europe, private by default
See original launch post

TL;DR: Munito is an open-model inference provider for Europe, private by default. Point your OpenAI SDK or Claude Code at Munito and run GLM, Qwen, DeepSeek and more, with inference and storage in the EU. No GPUs to manage.

https://youtu.be/5tjInE84mY4

Hi everyone! I'm Alexey, founder of Munito AI.

I led inference at Mistral AI, where my team built its API and inference products, and I built Le Chat's Deep Research. Before that, I spent five years at Airbnb, where I built Hosting Services from scratch and grew it into a global business.

At Mistral AI, I kept seeing the same thing: Europe lacks well-established inference providers for open models at scale. Cloud providers sell inference alongside their VMs, and model labs sell it for their own models, often not compatible with the major SDKs or agents.

I believe that open and independent intelligence, both models and inference, is key to economic growth and prosperity in the world and in Europe.

This is why I founded Munito: the best open models, run in Europe, behind an API you already know how to use.

What you get

  • Your existing code. OpenAI- and Anthropic-compatible APIs. Change the base URL and the key, and keep your SDK and agent tools. Tool calling works.
  • EU inference and storage. On the EU endpoint, https://api.eu.munito.ai, requests are processed in the EU, and what you store stays there by default.
  • A US region on the same stack. Need your inference processed in the US? Point the same code at https://api.us.munito.ai.
  • Private by default. Zero data retention by default. We never sell your data or train on it.
  • Text, speech and documents in one API. Open LLMs, speech to text and document AI, with one key.
  • A free tier for everyone. Sign up and start building, with no sales call. Buy credits for higher limits, or talk to us about monthly invoicing for your company.

Models: GLM-5.3-Flash, Qwen3.8 and DeepSeek-V4.1-Flash for text and agents, Kolibri-1 from Aleph Alpha, LightOnOCR-2 from LightOn for documents, and Gemma 4 for speech to text, all served in the EU. Need a model we don't serve yet? Tell us at founders@munito.ai, and we'll add it.

Who it's for: companies building with AI whose users care where their data goes, from a first agent to a product used by teams in finance, health and law.

What's next: we're building a self-improving inference stack, so running these open models gets faster and cheaper over time.

How to start

  1. Try it now on the free tier: sign up at munito.ai/registration, with no sales call. Any feedback is welcome!
  2. Building on open models? Email founders@munito.ai with your model and use case, and we'll help you get it running on Munito.
  3. Selling to European customers who keep asking where their AI runs? Talk to an engineer.

Merci! Alexey

YC Photos
Hear from the founders

How did your company get started? (i.e., How did the founders meet? How did you come up with the idea? How did you decide to be a founder?)

I led inference at Mistral AI, where my team built its API and inference products, and I built Le Chat's Deep Research. Before that, I spent five years at Airbnb, where I built Hosting Services from scratch and grew it to become a global product.

At Mistral, we built everything in-house, because no independent European provider could handle our scale. I kept seeing the same thing: in Europe, inference is a side product. Cloud providers sell it next to their VMs, and model labs sell it for their own models. At the same time, open models became really good.

This is why I started Munito: the best open models, run in Europe, behind an API you already know how to use.

What is the core problem you are solving? Why is this a big problem? What made you decide to work on it?

Companies in Europe want to use the best open models at scale, but they also need privacy and control over where their data is processed. Today it's hard to get both. Most large providers are not European, and EU processing is quite often an extra option instead of the default. There are few European providers, and many of them can't handle real scale.

And this problem grows every month. More and more sensitive data goes through AI models in products, agents and internal tools. Users and regulators ask where it's processed and stored, and companies don't want their AI to depend on one cloud or one model lab.

I saw this from inside at Mistral, where we had to build everything in-house. Munito gives companies open models at scale, with inference and storage in the EU and privacy by default: no data retention, no training on their data. Sovereignty, without running your own GPUs.

What is your long-term vision? If you truly succeed, what will be different about the world?

I believe that open and independent intelligence, both models and inference, is key to economic growth and prosperity in the world and the region.

If Munito succeeds, anyone building with AI can run the best open models privately, in the region they choose, without depending on one cloud or one model lab. Munito rents GPUs across many vetted data centers and clouds, and we’re building a self-improving inference stack, so running open models gets faster and cheaper over time.

Munito AI
Founded:2026
Batch:Fall 2026
Team Size:1
Status:
Active
Location:Paris, France
Primary Partner:Aaron Epstein