← Back to home
Model catalog
Every model, one endpoint.
All 63 models reachable through the same OpenAI-compatible endpoint, grouped by provider. Each row shows the EU territorial tiers it can be routed within and its current availability.
| Model | Provider | Size | EU-Sweden | EU-Strict | EU-Cloud | Unrestricted | Status |
|---|---|---|---|---|---|---|---|
 Mistral8 models | |||||||
| Mistral DevstralEU | Mistral | L | Operational | ||||
| Mistral LargeEU | Mistral | L | Operational | ||||
| Magistral MediumEU | Mistral | M | Operational | ||||
| Mistral CodestralEU | Mistral | M | Operational | ||||
| Mistral MediumEU | Mistral | M | Operational | ||||
| Magistral SmallEU | Mistral | S | Operational | ||||
| Mistral Pixtral 12BEU | Mistral | S | Operational | ||||
| Mistral SmallEU | Mistral | S | Operational | ||||
 Evroc7 models | |||||||
| Mistral Medium 3.5 (Evroc EU)EU | Evroc | M | Operational | ||||
| GLM-5.2 (Evroc EU)EU | Evroc | L | Operational | ||||
| GPT-OSS 120B (Evroc EU)EU | Evroc | L | Operational | ||||
| Kimi K2.6 (Evroc EU)EU | Evroc | L | Operational | ||||
| Llama 3.3 70B (Evroc EU)EU | Evroc | L | Operational | ||||
| Qwen3-VL 30B (Evroc EU)EU | Evroc | L | Operational | ||||
| Qwen3.6 35B (Evroc EU)EU | Evroc | M | Operational | ||||
| Anthropic12 models | |||||||
| Claude Fable 5 | Anthropic | L | Operational | ||||
| Claude OpusEU | Anthropic | L | Operational | ||||
| Claude Opus 4.5EU | Anthropic | L | Operational | ||||
| Claude Opus 4.6EU | Anthropic | L | Operational | ||||
| Claude Opus 4.7 | Anthropic | L | Operational | ||||
| Claude Opus 4.8 | Anthropic | L | Operational | ||||
| Claude SonnetEU | Anthropic | M | Operational | ||||
| Claude Sonnet 4.5EU | Anthropic | M | Operational | ||||
| Claude Sonnet 4.6EU | Anthropic | M | Operational | ||||
| Claude Sonnet 5 | Anthropic | M | Operational | ||||
| Claude HaikuEU | Anthropic | S | Operational | ||||
| Claude Haiku 4.5EU | Anthropic | S | Operational | ||||
| OpenAI12 models | |||||||
| GPT-5 | OpenAI | L | Operational | ||||
| GPT-5.4 (Azure EU)EU | OpenAI | L | Operational | ||||
| GPT-5.5 | OpenAI | L | Operational | ||||
| GPT-o3 | OpenAI | L | Operational | ||||
| GPT-4.1 | OpenAI | M | Operational | ||||
| GPT-4o | OpenAI | M | Operational | ||||
| GPT-5 Mini | OpenAI | M | Operational | ||||
| GPT-o4 Mini | OpenAI | M | Operational | ||||
| GPT-4.1 Mini | OpenAI | S | Operational | ||||
| GPT-4.1 Nano | OpenAI | S | Operational | ||||
| GPT-4o Mini | OpenAI | S | Operational | ||||
| GPT-5 Nano | OpenAI | S | Operational | ||||
| berget7 models | |||||||
| GLM-4.7 (Berget EU)EU | berget | L | Operational | ||||
| GPT-OSS 120B (Berget EU)EU | berget | L | Operational | ||||
| Kimi K2.6 (Berget EU)EU | berget | L | Operational | ||||
| Llama 3.3 70B (Berget EU)EU | berget | L | Operational | ||||
| Gemma 4 31B (Berget EU)EU | berget | M | Operational | ||||
| Mistral Medium 3.5 (Berget EU)EU | berget | M | Operational | ||||
| Mistral Small 3.2 (Berget EU)EU | berget | S | Operational | ||||
 Amazon1 model | |||||||
| Nova LiteEU | Amazon | M | Operational | ||||
| Nebius6 models | |||||||
| GPT-OSS 120B (Nebius) | Nebius | L | Operational | ||||
| Llama 3.3 70B (Nebius) | Nebius | L | Operational | ||||
| Qwen2.5-VL 72B (Nebius) | Nebius | L | Operational | ||||
| Qwen3 235B (Nebius) | Nebius | L | Operational | ||||
| Gemma 3 27B (Nebius) | Nebius | M | Operational | ||||
| Qwen3 32B (Nebius) | Nebius | M | Operational | ||||
| xAI5 models | |||||||
| Grok 4 | xAI | L | Operational | ||||
| Grok 4.3 | xAI | L | Operational | ||||
| Grok 4.5 | xAI | L | Operational | ||||
| Grok 4 Fast | xAI | M | Operational | ||||
| Grok 3 Mini | xAI | S | Operational | ||||
 Google4 models | |||||||
| Gemini 3.1 Pro | L | Operational | |||||
| Gemini 3.5 Flash | M | Operational | |||||
| Gemini Pro | M | Operational | |||||
| Gemini Flash | S | Operational | |||||
| Z.ai1 model | |||||||
| GLM-5.2 | Z.ai | L | Operational | ||||
Availability reflects the published catalog. Live per-request routing always honours your team's territorial tier as a hard constraint — a model is only served where it qualifies.