Lyceum ๐ฉ๐ช
EU-hosted inference cloud for open-source models, OpenAI-compatible
From $0.06/1M tokens
- Featured
- GDPR
Tools whose operating company is itself under EU, EEA, UK or Swiss law, verified from each company's own legal pages.
Keeping the processor inside the EU means no transfer mechanism to argue about, which is where EU-US transfers have repeatedly come unstuck.
EU-headquartered companies are not subject to the US CLOUD Act, which can compel US-based providers to hand over data regardless of where it is stored.
The EU AI Act requires logging, monitoring and audit trails for high-risk AI systems. Those obligations have applied since 2 August 2026. The classification rules in Article 6(1) follow on 2 August 2027, so the Act is still phasing in.
German public sector requires BSI C5; French healthcare mandates HDS certification with EEA-only hosting.
HQ country comes from each company's imprint or terms, never from hosting location or domain. GDPR, SOC 2 and HIPAA badges appear only where the vendor states them; a missing badge means unverified, not absent. Corrections go through the edit link on every product page.
No tools match these filters.
Nothing in this directory carries every attribute at once. Widen it by one:
ยท Missing a European product? Submit it.
Lyceum ๐ฉ๐ช
EU-hosted inference cloud for open-source models, OpenAI-compatible
From $0.06/1M tokens
Akumi ๐ณ๐ฑ
EU-hosted OpenAI-compatible inference API with data residency, PII pseudonymization and per-request audit trails
Pay-as-you-go, EUR 10 minimum top-up
Cortecs AI ๐ฆ๐น
European AI inference gateway with smart routing across EU providers
Pay-per-use + 5% gateway fee
EUrouter ๐ณ๐ฑ
European AI gateway that routes to 100+ models with EU data residency
Free (10K req/mo), 39 EUR/mo (Plus)
GreenPT ๐ณ๐ฑ
French inference API for open-weight models, hosted on Scaleway with embeddings, reranking and speech
IONOS AI Model Hub ๐ฉ๐ช
OpenAI-compatible API for open-weight LLMs and image models, hosted in IONOS EU data centers
From $0.11/1M tokens (Qwen3.5-9B)
Infercom ๐ฑ๐บ
European sovereign AI inference with OpenAI-compatible APIs hosted in EU datacenters
From โฌ0.20/M input tokens (gemma-4-31B)
LLM Tech ๐ต๐ฑ
EU inference provider serving Qwen3.8-27B from dedicated Helsinki GPUs it rents and operates itself, with zero data retention
From $0.25/1M input tokens
LibertAI ๐ซ๐ท
Decentralized, privacy-first inference API running open-source LLMs in trusted execution environments
Melious AI ๐ฉ๐ช
European inference API serving 60+ open-weight models on OpenAI- and Anthropic-compatible endpoints
Mistral ๐ซ๐ท
Use models in a few clicks with our platform. Download our open models for deep access.
From $0.10/1M tokens
Nebius ๐ณ๐ฑ
Full-stack AI cloud with GPU infrastructure for training and inference
From $2.00/hr (H100)
Nscale ๐ฌ๐ง
European AI hyperscaler with serverless inference and GPU cloud
From $0.01/M tokens
Opper ๐ธ๐ช
EU-hosted AI gateway serving 300+ models through one OpenAI-compatible API
From Provider token rates + 3% credit fee
Scaleway ๐ซ๐ท
European serverless AI inference APIs, 100% hosted in Europe
From โฌ0.15/M input tokens (chat models)
SimpleLLM ๐ฉ๐ช
OpenAI-compatible API for open-weight models, hosted only in EU data centres
Solheim AI ๐ฎ๐น
Private EU-hosted LLM instances billed at a flat monthly fee rather than per token
From EUR 15/month
TensorX ๐ฎ๐ช
EU-sovereign inference API with 42+ open-source models and zero data retention
From $0.02/M tokens
WAYSCloud ๐ณ๐ด
Norwegian cloud platform with an LLM inference API running open-weight models in Norway
regolo ๐ฎ๐น
OpenAI-compatible inference API run on Italian infrastructure with zero data retention
AiQu ๐ธ๐ช
Swedish GPU infrastructure and LLM hosting platform with API-first deployment, no Kubernetes required
From 49 SEK/month
Airon ๐ธ๐ช
Dedicated bare-metal GPU infrastructure for AI workloads, hosted in Nordic datacenters
From $1.24/hour
Hyperstack ๐ฌ๐ง
On-demand cloud GPU platform for AI and ML workloads with per-minute billing
From $0.15/hr
OVHcloud AI ๐ซ๐ท
European cloud provider with AI inference, training, and deployment services
From $0.91/hr (L4 GPU)
Taiga Cloud ๐ฉ๐ช
European GPU cloud for AI training and inference by Northern Data Group
From ~โฌ2.70/GPU-hr
Verda ๐ซ๐ฎ
European GPU cloud with on-demand instances and serverless inference
From $0.14/hr
evroc ๐ธ๐ช
European-sovereign cloud and inference APIs running open-source models on NVIDIA Blackwell GPUs in EU data centers
Jina AI ๐ฉ๐ช
Search APIs for embeddings, reranking, and web-to-markdown conversion
Free (10M tokens)
Prem AI ๐จ๐ญ
Fine-tune and deploy LLMs on your own infrastructure with full data sovereignty
Varion ๐ฑ๐น
OpenAI-compatible proxy that trims LLM token usage before requests reach your provider
From EUR 19/month (10M tokens)
Agenta ๐ฉ๐ช
Open-source prompt management, evaluation, and observability for LLM apps
Free
Giskard ๐ซ๐ท
Eliminate risks of biases, performance issues & security holes in AI models. In <10 lines of code.
LangWatch ๐ณ๐ฑ
LLM observability platform with quality monitoring, guardrails, and evaluation workflows
Free / โฌ499/mo
Langfuse ๐ฉ๐ช
Traces, evals, prompt management and metrics to debug and improve your LLM application.
From $29/mo
Rhesis AI ๐ฉ๐ช
Open-source testing platform for LLM and agentic applications. Test generation, adversarial probing, and regression tracking.
Weckr ๐ธ๐ช
Tracks LLM cost and margin per end customer, with spending caps that fire before the call
Free up to 50k requests/mo
Dust ๐ซ๐ท
The operating system for AI agents
From โฌ29/user/mo
Langdock ๐ฉ๐ช
GDPR-compliant enterprise AI platform with multi-LLM access and agents
From โฌ20/user/mo
PromptLeo ๐ฉ๐ช
No-code AI digital employees (Support, SDR, Research, Ops) built on open agent models, EU-hosted
From $99/mo (early access, list price $500/mo) + your own LLM API costs
Pydantic AI ๐ฌ๐ง
Type-safe Python agent framework with Pydantic validation, tool calling, and dependency injection
Rasa ๐ฉ๐ช
Conversational AI platform for building enterprise AI agents
Free (Developer), $35,000/yr (Growth)
Samtal ๐ธ๐ช
Swedish-hosted voice AI API with TTS, ASR, voice cloning, and conversational agents. ElevenLabs-compatible
n8n ๐ฉ๐ช
AI workflow automation with 400+ integrations and native AI agents
From โฌ24/mo
Roofline ๐ฉ๐ช
Edge AI deployment stack: an MLIR/IREE compiler and device runtime for CPUs, GPUs and NPUs
Atomic Chat ๐ช๐ช
Open-source local AI chat app for running open-weight models on desktop and mobile
Pydantic AI ๐ฌ๐ง
Type-safe Python agent framework with Pydantic validation, tool calling, and dependency injection
llmkit ๐ซ๐ฎ
One LLM client API for 20+ providers, in Go, TypeScript, Python and Rust
Gladia ๐ซ๐ท
Fast speech-to-text API with real-time transcription and speaker diarization
From $0.61/hr
LemonFox ๐ฉ๐ช
Affordable speech-to-text and text-to-speech API with 100+ language support
From $5/mo
Samtal ๐ธ๐ช
Swedish-hosted voice AI API with TTS, ASR, voice cloning, and conversational agents. ElevenLabs-compatible
Speechmatics ๐ฌ๐ง
Enterprise speech-to-text API supporting 55+ languages with high accuracy
From $0.24/hr
Meilisearch ๐ซ๐ท
Open-source search engine in Rust with full-text, semantic, and hybrid search
Free (self-hosted), $20/mo cloud
Qdrant ๐ฉ๐ช
Powering the next generation of AI applications with advanced and high-performant vector similarity search technology.
From ~$25/mo
Vespa ๐ณ๐ด
Open-source search and serving engine combining vector search, keyword search, and ML ranking at scale
Free (self-hosted), cloud usage-based
Weaviate ๐ณ๐ฑ
Weaviate is an open source, AI-native vector database that helps developers create intuitive and reliable AI-powered applications.
From $45/mo
OVHcloud AI ๐ซ๐ท
European cloud provider with AI inference, training, and deployment services
From $0.91/hr (L4 GPU)
Prem AI ๐จ๐ญ
Fine-tune and deploy LLMs on your own infrastructure with full data sovereignty
15 of the inference APIs here are GDPR-compliant, headquartered in Europe, and run the models themselves instead of routing them elsewhere, 8 of them: Lyceum (Germany), Berget AI (Sweden), GreenPT (Netherlands), IONOS AI Model Hub (Germany), Infercom (Luxembourg), LLM Tech (Poland), Melious AI (Germany), and Mistral (France), and 7 more in the grid above. A European head office puts the company holding your account and your data under EU law rather than the CLOUD Act. It says nothing about which data centre serves a given request, so if you need residency and not only jurisdiction, the product page has it.
7 of the listings above are gateways: Akumi, Cortecs AI, EUrouter, Opper, Orq.ai, Tokenade, and Varion. Most put one API in front of many models, so you can swap models without rewriting your integration, and several add key management, fallbacks or caching on top. That flexibility is the usual reason to reach for one. Others are narrower proxies that sit in front of a model to cut token usage, so the taglines above are worth a read before you pick. Either way, the gateway's own jurisdiction covers your account and your logs rather than the model behind them, so a gateway can be European while a request it routes ends up at a US model lab. Where that distinction matters to you, each product page records which models the provider routes to and who runs them.
Major US cloud and AI providers also offer EU-region deployments. Data stays in European data centres, but the provider stays subject to US jurisdiction, including the CLOUD Act. That is usually fine. It stops being fine when a regulator, a customer contract or a sector rule asks who could be compelled to hand the data over. Regions as of August 2026.
AWS
EU regions in Frankfurt, Ireland, Stockholm, Paris, Milan, Spain, Zurich. Sovereign Cloud in Germany.
Microsoft Azure
EU regions in Netherlands, Ireland, France, Germany, Sweden, Switzerland, Norway, Poland, Italy, Spain.
Google Cloud
EU regions in Belgium, Netherlands, Finland, Germany, France, Poland, Italy, Spain. Sovereign Controls available.
61 AI infrastructure tools headquartered in Germany, France, Sweden, Netherlands, United Kingdom, Finland, Italy, Norway, Estonia, Austria, Luxembourg, Poland, Switzerland, Ireland, and Lithuania, across 8 categories. 5 of them appear in more than one category, so the section counts add up to more than 61. Compliance badges come from the vendor's own statements, so a missing badge means the claim was not found, not that the answer is no. For the full landscape including US-based tools, see the AI Infrastructure Landscape. Missing a European product? Submit it here.
Is your product missing?