EU options at a glance
| Provider | Free access | Where data is processed | Training on API data |
|---|---|---|---|
| OVHcloud AI Endpoints | Anonymous: 2 requests per minute per IP and model | "Hosted and run in the European Union" | "Your data is never used to train our models" |
| Scaleway Generative APIs | First 1,000,000 tokens free, then billed | A data center in Paris, France | Not used for training |
| Mistral | Free mode has no API access | EEA by default. The api.eu.mistral.ai endpoint keeps inference in EU and EFTA data centers. | Depends on the product default and your opt-out |
OVHcloud AI Endpoints
OVHcloud is a French cloud provider. Its product page says that AI Endpoints services are hosted and run in the European Union. The documentation says that data is not stored or shared during or after model use. OVHcloud keeps only the data that it needs for billing.
- Anonymous use: 2 requests per minute, for each IP address and each model. No account is needed.
- With a key: 400 requests per minute, for each Public Cloud project and each model. The project needs a payment method.
- Base URL:
https://oai.endpoints.kepler.ai.cloud.ovh.net/v1.
| Model | Free on | Context | Free limits (best offer) | Tools | Vision |
|---|---|---|---|---|---|
| Qwen3.8 27B | OVH AI Endpoints | 262K | 2 req/min 2/min per IP | Yes | No |
| Qwen3.6 27B | OVH AI Endpoints | 131K | 2 req/min 2/min per IP | Yes | No |
| Qwen3.5 397B | OVH AI Endpoints | 262K | 2 req/min 2/min per IP | Yes | No |
| Qwen3.5 9B | OVH AI Endpoints | 262K | 2 req/min 2/min per IP | Yes | No |
| GPT-OSS 120B | OVH AI Endpoints | 131K | 2 req/min 2/min per IP | Yes | No |
| GPT-OSS 20B | OVH AI Endpoints | 131K | 2 req/min 2/min per IP | Yes | No |
| Llama 3.3 70B | OVH AI Endpoints | 131K | 2 req/min 2/min per IP | Yes | No |
| Mistral Small 3.2 24B | OVH AI Endpoints | 131K | 2 req/min 2/min per IP | Yes | No |
Showing 8 of 11 matching free models. See all models.
Live data from the FreeLLMAPI catalog (version 2026.11.19, checked 19 November 2026). Free limits change often. The catalog is checked every day.
Scaleway Generative APIs
Scaleway is also a French cloud provider. Its FAQ says that all Serverless models are hosted in a data center in Paris, France. Scaleway says that models on api.scaleway.ai will stay hosted in Europe. It will tell customers before the location changes from France to other European regions.
The free tier covers the first 1,000,000 tokens and 60 minutes of audio transcription. After that, Scaleway charges per million tokens. The data privacy page says that your data is not used for training and is not accessible to the creators of the models. Aggregated and anonymized usage data is kept for up to 6 months.
Scaleway is not in the FreeLLMAPI free catalog, because its free tier is a one-time token amount. You can add it as a custom provider. Read the last section of this page.
Mistral
Mistral is a French AI company. Its help center says that data is hosted in the European Economic Area by default. Some features can move data outside the EU for a short time. Mistral lists those locations in its Trust Center.
The regional inference page gives more detail. The global endpoint api.mistral.ai is the default. For that endpoint, Mistral "does not commit to a specific inference location". The EU endpoint api.eu.mistral.ai runs inference in data centers in EU and EFTA countries. It costs more, and Agents, Batch and the Files API are not available on it.
Gemini and other providers for EU users
- Google Gemini: if your app is available to users in the EEA, Switzerland or the UK, the terms allow only Paid Services. For users in those regions, the Paid Services data terms apply to all services, including unpaid quota. Google does not use that content to improve its products.
- Groq: Groq says that all customer data is retained in Google Cloud buckets in the United States. Standard contractual clauses can apply to transfers.
- Cloudflare Workers AI: Cloudflare does not use Workers AI content to train models. We did not check the processing location.
Keep requests on EU providers with FreeLLMAPI
FreeLLMAPI is an open-source gateway that runs on your own computer or server. You choose which providers it calls. Your keys stay in a local encrypted database. The FreeLLMAPI catalog server never sees your prompts.
- Install FreeLLMAPI on a computer or server that you control.
- Open the Keys page. Add a custom provider for Scaleway with the base URL
https://api.scaleway.ai/v1and your Scaleway key. - On the Fallback page, create a chain with the name
eu. Keep the default "Start empty". - Add only OVHcloud and Scaleway models to the
euchain. - Send requests with the model
auto:eu. - Check the
X-Routed-Viaheader to confirm the provider.
Request through the EU-only chain
curl http://localhost:3001/v1/chat/completions \
-H "Authorization: Bearer <unified-key>" \
-H "Content-Type: application/json" \
-d '{"model": "auto:eu", "messages": [{"role": "user", "content": "Bonjour"}]}'A named chain uses only the models in it. If the chain has no enabled models, the router returns error 400. It does not fall back to the full catalog. The router does not pin regions by itself, so the chain is how you keep requests in the EU.
Frequently asked questions
Is there a free LLM API hosted in the EU?
Yes. OVHcloud AI Endpoints is hosted and run in the EU and allows anonymous use at 2 requests per minute per IP and model. Scaleway gives the first 1,000,000 tokens free from a data center in Paris.
Is the Mistral API free and GDPR compliant?
Mistral hosts data in the EEA by default, and its terms include a Data Processing Agreement. Free mode does not include API access, so the API needs billing activation.
Can I use the free Gemini API for users in Europe?
Google's terms say that apps available to users in the EEA, Switzerland or the UK must use Paid Services. For those users, Google does not use content to improve its products.
Does OVHcloud AI Endpoints store my prompts?
OVHcloud says that data is not stored or shared during or after model use, and it is never used to train models. It keeps only the data required for billing.
Related guides
- Do free LLM APIs train on your data?
- LLM API without signup
- OVH AI Endpoints free API
- Free LLM API for commercial use
- Mistral API free tier: what you get in 2026
Sources
- OVHcloud AI Endpoints (checked 20 November 2026)
- OVHcloud: AI Endpoints capabilities (checked 20 November 2026)
- OVHcloud: AI Endpoints getting started (checked 20 November 2026)
- Scaleway Generative APIs: FAQ (checked 20 November 2026)
- Scaleway Generative APIs: Data privacy (checked 20 November 2026)
- Mistral help: Where do you store my data (checked 20 November 2026)
- Mistral docs: Regional inference (checked 20 November 2026)
- Mistral docs: Usage and limits (checked 20 November 2026)
- Mistral Commercial Terms of Service (checked 20 November 2026)
- Gemini API Additional Terms of Service (checked 20 November 2026)
- GroqCloud: Your data (checked 20 November 2026)
- FreeLLMAPI docs: Named fallback chains (checked 20 November 2026)