> Source: https://souverana.ch/insights/souveraene-ki-infrastruktur-en/ > Language: en # Sovereign AI infrastructure: why the model is the easy part The language model is the easiest part of your AI to swap. What ties you to one provider are the models that read your documents and make them searchable (OCR and embeddings). Which Swiss hosters offer every component. Guide · Published 14 Jul 2026 · Updated 21 Oct 2026 · Joel Barmettler ## What infrastructure actually means for AI **You can swap the language model of your AI in an afternoon. Switch the embedding model, which translates your texts into series of numbers for search, and you have to re-ingest all your documents. When AI is meant to answer questions from your own documents, it needs further services: OCR (text recognition) reads scans as text, transcription writes out audio.** **In brief** - Compare providers by how much of the data preparation they take on for you: ingesting documents, preparing them for search and storing them. The more the provider takes on, the harder a later switch becomes. That is why the most convenient offer ties you in most tightly. - For searching your documents (embedding) and for reading scans (OCR), use open models that you can also operate yourself. Only then can you switch provider later. - Choose a hoster that runs all the data-preparation models in Europe. If a hoster offers only language models, you have to fall back on a US API for embeddings and OCR. As of July 2026, only PHOENIQS/kvant, a hoster for open AI models, offers the complete set in Switzerland. - For the language model, performance and running costs are enough to go on, because that is the easiest decision to correct. Public debate revolves almost entirely around language models, meaning AI like ChatGPT that understands and writes text. For a company, though, the language model is only the final step in a longer chain. This is the third part of our series on sovereign AI. It covers the lowest of three layers, where your data is prepared. **Series · Sovereign AI** 1. [What sovereign AI actually means for Swiss companies](https://souverana.ch/en/insights/was-ist-souveraene-ki/) 2. [AI strategy for SMEs: the questions to answer first](https://souverana.ch/en/insights/ki-strategie-kmu/) 3. Sovereign AI infrastructure: why the model is the easy part (this part) 4. [Sovereign AI platform: why it binds more than the model](https://souverana.ch/en/insights/souveraene-ki-plattform/) 5. [Sovereign AI integration: when AI starts to act](https://souverana.ch/en/insights/souveraene-ki-integration/) ## The model does not know your documents A language model is trained on the public internet. It knows nothing about your company: it has never seen your contracts, quotes or meeting minutes. Yet anyone using AI in a business wants exactly that: answers from their own documents. Those documents exist as PDFs, office files, scanned contracts and recordings of meetings. Before a language model can use them, they have to be turned into searchable text. The rest of the infrastructure handles that: an OCR model makes documents and scans readable, transcription writes out audio, an embedding model makes the text searchable, a vector database stores it, and only at the end does the language model compose an answer from the matching passages. This interplay is called RAG, Retrieval-Augmented Generation, and it runs in two phases. Once, you ingest your documents and store them in searchable form. After that, every question runs through the second phase, from the search through the matching passages to the evidenced answer. A RAG pipeline as a graph. At the top, your knowledge is prepared once and stored in the vector database; at the bottom, every question runs through the search to an evidenced answer. The database connects both branches, and the embedding model is the same one in both. Four of the stations are AI models in their own right, and the language model at the end is only one of them. Each of these stations is its own decision, with its own providers and its own consequences for sovereignty. Anyone who only looks at the language model overlooks most of the chain. We therefore start with the question of how a company obtains AI in the first place. ## How companies obtain AI How sovereign a solution becomes depends above all on how much work the provider takes off your hands. Where the provider works for you, it also keeps control afterwards. Four paths, from the greatest convenience to the greatest control: | Path | Provider | What you get | Data preparation | | --- | --- | --- | --- | | Suite with AI over your data | Microsoft, Google, Infomaniak | ready-made apps plus AI on top | the provider, in its own storage | | AI platform with ingestion | OpenAI, Mistral | components plus managed storage | the provider, in its cloud | | Cloud with model APIs | OVHcloud | models as a service, you build the pipeline | you | | Open models, self-operated | Scaleway, PHOENIQS, on-premise | open models, full control | you | The most convenient are the **suites**. Microsoft and Google dominate enterprise AI because your documents and mail already sit with them. [Copilot searches SharePoint and OneDrive](https://souverana.ch/en/insights/microsoft-copilot-datenschutz/), Gemini searches Google Drive; the AI reads the data you already keep there anyway. Infomaniak belongs in the same category, with Swiss infrastructure: Euria searches the files in your kDrive and runs on open models in Switzerland. Infomaniak is a hosting provider with built-in AI, not a company specialised in AI. The next category is the **AI platforms**, whose real product is the AI. OpenAI, and Mistral as the European answer, give you the components through an interface and offer to upload, ingest and make your documents searchable. Mistral covers the full range of this: its own assistant (Le Chat), a platform (AI Studio), OCR, transcription, embeddings and on-premise operation. A sovereign provider can therefore keep up with the US corporations, from the assistant to the interface. Next come the **cloud providers** such as OVHcloud, whose core business is infrastructure and who offer AI models as an additional service. The final step is to run open models only: at Scaleway, at a Swiss provider such as PHOENIQS, or on your own hardware. The more self-contained the operation, the more you set up yourself, and the less a provider can later take away from you. ## Where the lock-in actually sits The language model is the most easily replaceable component. The instruction to the model, the prompt, stays the same, and the same open model runs at a dozen providers. The dependency arises at the models that process your data. For search and RAG, you convert your entire knowledge base into vectors with one specific embedding model, which turns each passage of text into a series of numbers representing its meaning. Documents and search queries must use the same model, or the vectors will not match. The embedding model therefore determines how your entire data set is built. Switch it, and you have to **re-ingest and re-index every document**, because one model’s vectors are meaningless to another. Highlighted: the embedding model. It appears twice, at indexing and at every question, and must be the same both times. A switch devalues the entire vector database. For this reason, OpenAI, Google and Mistral offer to ingest and store your documents for you. At OpenAI, the resulting knowledge base then sits in their vector database, billed per gigabyte and day. At Google, the RAG engine handles ingestion, chunking and embedding. It is convenient, and this is exactly where the strongest dependency arises: the ingested, embedded and stored data set sits in one provider’s vector space and storage, and switching means rebuilding it in full. The second difficult component is the preparation itself. Extracting clean, structured data from real documents (Swiss PDFs, scans, tables and three national languages) is the real engineering work. The pipeline behind it is tailored to one specific OCR service’s output format and is just as hard to replace afterwards. Highlighted: OCR and transcription at the start of the chain. The OCR output format shapes everything that follows and locks you into one provider. At this layer, sovereignty therefore mainly means keeping the models that process your data, namely embeddings and OCR, open and, if necessary, able to be run by yourself. ## Choosing the model That leaves the question most people ask first: which language model? It matters, but it is the easiest decision to correct. Two reference points are enough to start. For performance, the independent Intelligence Index from Artificial Analysis ranks models by capability on standardised tests, open and closed models together. For cost, what we calculated in our article on [LLM cost](https://souverana.ch/en/insights/llm-kosten/) applies: the cheapest model on the price list is rarely the cheapest one in operation, because token consumption per task varies from model to model. What matters for sovereignty is that the model has open weights, meaning its trained parameters are published. Then you can run it at several hosters, compare their prices and switch whenever you want. ## Choosing the hoster Anyone who wants to be sovereign chooses a hoster that runs open models. What matters is whether the hoster offers all the models that process your data on European or Swiss soil. If it only runs language models and you have to fall back on a US API for embeddings or OCR, sovereignty is lost again at exactly the point that matters most. > **six model classes** A single Swiss provider, PHOENIQS/kvant, today operates language models, OCR, embeddings, a reranker (ranks search hits by relevance), transcription and vision (image understanding), all as open models on Swiss servers (among others Apertus 70B, MinerU, BGE-M3, Whisper). > > Source: PHOENIQS Model-as-a-Service, catalogue as of July 2026 Four model types sit inside the pipeline (in red): OCR and transcription, embedding, reranking, language model. A sovereign hoster has to offer all four, including the models before the language model. European providers cover these model types to differing extents. OVHcloud lists over forty open models including transcription, embeddings and image models. Scaleway operates language models, vision and embeddings on sovereign infrastructure in Paris. Infomaniak offers the same models through an OpenAI-compatible API from Switzerland, so existing software can easily be switched over. PHOENIQS shows that a Swiss provider can run all the models for data preparation openly, OCR and embeddings included.[1] The greatest control comes from running open models yourself on your own or a rented GPU (graphics processor). That is a topic in its own right, [which we cover in a dedicated article](https://souverana.ch/en/insights/llm-selbst-hosten/): the setup is demanding, but ongoing operation afterwards requires little effort. ## Do not decide in isolation One more note on order: the models depend on the platform, that is, the software in which your team works with the AI. The platform requires a particular interface and often a matching embedding model, and so dictates part of the infrastructure. That is why we stay with the model and data preparation here, and cover the platform in the next article of this series. ## What this means for your company Sovereign infrastructure delivers AI that is just as good. You assemble the individual components yourself instead of renting a finished, closed solution. The language model is the easy part. Your care belongs with the models that process your data, and with the question of whether you could take your data and models to another provider if necessary. Want model, hoster and data preparation chosen independently for your case? [Architecture & Solution Selection](https://souverana.ch/en/leistungen/ki-architektur/) Above the infrastructure sits the layer your team works with every day. The next part of the series is devoted to that layer, and the last part turns to the integration layer. ## Sources [1] PHOENIQS. «PHOENIQS Model-as-a-Service, Katalog Stand Juli 2026». PHOENIQS. July 2026. Herstellerangabe. The author Joel Barmettler AI Architect · Souverana, Zurich Joel Barmettler guides Swiss companies from AI strategy to integration: sovereign, confidential and production-ready. He built the Swiss AI Hub as its architect and today co-owns its architecture; he personally leads every Souverana mandate. Mandates from one-person firms to Fortune 500 corporations. [Book an intro call](https://meet.brevo.com/joel-barmettler/30-minute-meeting) [More about Souverana](https://souverana.ch/en/) [LinkedIn](https://www.linkedin.com/in/joel-barmettler-b9ab361b7) ## Frequently asked questions What belongs to AI infrastructure? More than a language model. AI infrastructure also includes the models and services that prepare your data: OCR (text recognition) for documents, transcription for audio, embeddings (numerical vectors for search), and a layer for cost control and access. The language model is the most easily replaceable component of the set. Which AI model should a company choose? The model that solves the task at the best price. The Intelligence Index from Artificial Analysis, an independent comparison service, ranks models by capability. Running costs depend on token consumption (the amount of text you are billed for). An open model with freely available weights can be run at several hosters, and you can switch hoster at any time. Where can I host LLMs in Switzerland or Europe? As an API (programming interface) at European providers such as OVHcloud, Scaleway or Infomaniak. Or on a Swiss platform that offers models as a service, such as PHOENIQS/kvant, with open models including OCR and embeddings. Maximum control comes from running the models yourself, on your own or rented hardware. What is the difference between an AI platform and an integration? An integration is AI built into a product you already use (Microsoft Copilot, Google Gemini, Infomaniak Euria). The provider chooses the infrastructure behind it. A platform gives you the components through an interface, so you can build your own solution and switch provider. Why are embeddings a lock-in risk? An embedding model translates your texts into vectors, series of numbers that represent meaning, and your entire knowledge base is stored in that form. Documents and search queries must use the same model. Switching means re-ingesting and re-indexing every document. With an open, self-operated model, you decide when that rebuild happens. ## Glossary - **Embedding**: Converting a text into a list of numbers (a vector) that captures its meaning. Search and RAG compare these vectors. Important: documents and search queries must be converted with the same model, or the vectors will not match. - **OCR**: Optical Character Recognition: converting documents and scans (PDFs, images) into machine-readable, structured text. Often the first and hardest step in making your own documents usable for AI. - **API**: A programming interface through which software addresses a language model directly, without a chat interface. Usually billed per token. - **RAG**: Retrieval-Augmented Generation: before answering, the model receives matching excerpts from your own documents and answers with evidence instead of from memory. - **Prompt**: The input to a language model: question, instruction and any text sent along. What is in the prompt determines the answer and the cost. - **Token**: The smallest billing unit of a language model; a word is roughly one to three tokens. Providers charge per million tokens, separately for input and output. - **Open weight**: Models whose weights are publicly available and may be self-hosted (e.g. Apertus, DeepSeek, Llama). A precondition for sovereign operation without vendor lock-in. - **Parameters**: The learned numerical values of a language model. Their number describes its size: "8B" means eight billion parameters. Larger models can do more but need more GPU memory. - **GPU**: Graphics processing unit. Language models run almost exclusively on GPUs because they execute very many computations in parallel.