Skip to content
EU Hosting

AI Models

Claude, GPT, and Gemini, every model you use in Lurus Code runs in European data centers. Pick the right model per task without processing data outside the EU.

Anthropic

AWS Bedrock EU 🇪🇺

Claude Opus 4.6

Powerful
opus
Tool Calling Vision Adaptive Thinking Prompt Caching

Claude Sonnet 4.6

Balanced
sonnet
Tool Calling Vision Adaptive Thinking Prompt Caching

Claude Opus 4.5

Powerful
opus-4.5
Tool Calling Vision Extended Thinking Prompt Caching

Claude Sonnet 4.5

Balanced
sonnet-4.5
Tool Calling Vision Extended Thinking Prompt Caching

Claude Haiku 4.5

Fast
haiku
Tool Calling Vision Prompt Caching

Google

Vertex AI EU 🇪🇺

Gemini 2.5 Pro

Powerful
gemini-pro
Tool Calling Vision Reasoning Prompt Caching

Gemini 2.5 Flash

Fast
gemini
Tool Calling Vision Prompt Caching

OpenAI

Azure OpenAI EU 🇪🇺

GPT-5.4 Pro

Powerful
gpt-pro
Tool Calling Vision Reasoning Prompt Caching

GPT-5.4

Powerful
gpt
Tool Calling Vision Prompt Caching

GPT-5.3 Codex

Powerful
gpt-codex
Tool Calling Vision Prompt Caching

GPT-5.1 Codex Max

Powerful
gpt-codex-max
Tool Calling Vision Reasoning Prompt Caching

o4-mini

Powerful
o4-mini
Tool Calling Reasoning Prompt Caching

GPT-5.4 Mini

Fast
gpt-mini
Tool Calling Vision Prompt Caching

Open Source

Inference API EU 🇪🇺

GLM 5

Powerful
glm
Tool Calling

Kimi K2.5

Powerful
kimi
Tool Calling

Qwen3.5 397B

Balanced
qwen
Tool Calling

DeepSeek V3.2

Balanced
deepseek
Tool Calling

MiniMax M2.5

Balanced
minimax
Tool Calling

GPT OSS 120B

Fast
gpt-oss
Tool Calling

Capabilities, Legend

The tags on each card describe which capabilities that model supports in Lurus Code.

  • Tool Calling

    Structured tool and function calls, e.g. files, terminal, and integrations.

  • Vision

    Images and screenshots can be analyzed together with text.

  • Adaptive Thinking

    Scales reasoning depth dynamically based on task complexity (Claude adaptive thinking).

  • Extended Thinking

    Extra reasoning time for harder tasks (Claude extended thinking).

  • Reasoning

    Geared toward deeper multi-step and logical reasoning.

  • Prompt Caching

    Reuses prompt prefixes for lower latency and cost.

Automatic fallback chains

When a provider is temporarily unavailable, Lurus Code can automatically switch to an alternative model in the configured fallback chain, so your session keeps going without manual switching.

Ready to build in the EU?

Join the waitlist and use every model with EU hosting in one surface.

Get Started