Lurus combines EU-provider models, European models and frontier models from Anthropic, Google and OpenAI in one platform – unified across chat, agents and assistants.
Direct answer
All standard models are integrated into the platform. Zero Data Retention and No Training are required for every offered model; customer data is not used for model training.
Model status: 2026-07-19.
Models or operating paths with a European provider, hosting or privacy focus. Concrete provider information, No Training and Zero Data Retention matter most.
Lurus uses plan and token allowances. Powerful frontier models, long files and complex analyses may consume more allowance.
The best choice depends on privacy needs, context length, model portfolio, cost and workflow. The guide maps the criteria.
Also compare providers →Claude models for demanding analyses, code, long documents, and high-quality assistance in productive business processes.
Anthropic
Claude Opus 5 – powerful premium reasoning model with adaptive thinking, prompt caching, and tool-calling.
Anthropic
Claude Opus 4.8 – premium reasoning model with prompt caching and tool-calling.
Anthropic
Claude Opus 4.6 – premium reasoning model with prompt caching and tool-calling.
Anthropic
Claude Sonnet 5 – balanced reasoning model with adaptive thinking, prompt caching, and tool-calling.
Anthropic
Claude Haiku 4.5 – fast model with prompt caching and tool-calling.
Gemini models for fast answers, large context windows, multimodal analysis, and document-heavy workflows.
Gemini 3.7 Flash – fast Frontier model with vision and tool-calling.
Gemini 3.1 Flash Lite – efficient Frontier model with vision and tool-calling.
GPT models for versatile business tasks, complex reasoning, tool usage, and specialized coding workflows.
OpenAI
GPT-5.6 Sol – most capable OpenAI model for demanding reasoning, vision, and tool-calling.
OpenAI
GPT-5.6 Terra – balanced OpenAI model for strong reasoning, vision, and tool-calling.
OpenAI
GPT-5.6 Luna – efficient OpenAI model for fast reasoning, vision, and tool-calling.
OpenAI
GPT-5.5 – premium OpenAI model with prompt caching and tool-calling.
OpenAI
Versatile GPT model for writing, strategy, tool usage, and productive everyday assistance.
OpenAI
GPT-5.4 Mini – fast OpenAI model with prompt caching and tool-calling.
OpenAI
GPT-5.3 Codex – OpenAI coding model with prompt caching and tool-calling.
Models for privacy-oriented workflows, strong reasoning tasks, coding, and long contexts – with a focus on European provider and hosting options.
Moonshot AI
Kimi K2.7 Code – powerful EU-provider model with vision, tool calling, and strong reasoning.
Moonshot AI
Kimi K3 – built for complex tasks involving images, long documents, and multi-step tool workflows.
Zhipu AI
GLM 5.3 – suited for in-depth analysis, structured problem-solving, and reliable tool use.
Alibaba Cloud
Qwen 3.8 2.4T – excels at large text collections, precise analysis, and demanding knowledge tasks.
Zhipu AI
GLM 5.3 Flash – a fast, cost-efficient choice for everyday tasks, image understanding, and tool use.
DeepSeek
DeepSeek V4 Flash – a fast and efficient choice for text analysis, summaries, and tool-assisted tasks.
DeepSeek
DeepSeek V4 Pro – powerful EU-provider reasoning model for complex tasks.
Mistral AI
Mistral Medium 3.5 – balanced chat model with strong tool-calling.
Alibaba Cloud
Qwen 3.5 397B – capable chat model with vision and tool-calling.
Specialist models for image generation, Deep Research reports, and document/file analysis are used automatically depending on the task.
OpenAI
Specialist model for high-quality image generation directly from tasks, briefs, and prompts.
Black Forest Labs
Image model for maximum quality, precise visual execution, and professional results.
Lurus Routing
Lurus combines suitable EU-provider models for multi-step research, source analysis, and structured reports.
Lurus does not force a blanket choice between privacy and performance. Teams get a curated model portfolio with transparent provider choice and task-aware routing.
EU-provider models, Claude, Gemini, and GPT are standard parts of the platform and are used according to the task.
Zero Data Retention and No Training apply to every offered model. Customer data is not used for model training; providers or models without these commitments are not offered.
Depending on the request, Lurus can switch between fast, efficient, multimodal, and especially powerful models.
Lurus clearly communicates which model groups are available. Zero Data Retention, No Training, and EU Hosted are required for every offered model.
For tasks with a strong privacy focus, EU-provider models can be preferred. Hosted-in-Europe statements refer to this model group.
Claude, Gemini, and GPT are not a special “on request” list, but part of the curated standard portfolio.
These commitments apply to every offered model. If a provider or model does not support Zero Data Retention and No Training, it is not included in the portfolio.
The offered standard models are provided through EU-hosted operating paths. Models that do not meet these privacy and hosting requirements are not offered.
Lurus automatically selects the right model for your request. This overview shows typical use cases and suitable model groups.
| Use case | Recommended models | Why? |
|---|---|---|
| Marketing & copywriting | GPT-5.6 Terra · Claude Sonnet 5 · Kimi K2.7 Code | Strong writing quality, clear structure, and flexible tone for campaigns, newsletters, website copy, and social media. |
| Code & software development | GPT-5.6 Sol · GPT-5.3 Codex · Claude Sonnet 5 · Kimi K2.7 Code | Specialized coding and reasoning capabilities for reviews, architecture, debugging, and implementation. |
| Complex analyses & reasoning | Claude Opus 5 · GPT-5.6 Sol · GPT-5.6 Terra · Claude Opus 4.8 · GLM 5.3 · Qwen 3.8 2.4T | High reasoning quality and robust handling of long, multi-step tasks. |
| Quick everyday questions | GLM 5.3 Flash · DeepSeek V4 Flash · Gemini 3.1 Flash Lite · GPT-5.6 Luna | Fast answers for summaries, emails, short research, and internal assistance. |
| Data protection and EU focus | Kimi K3 · GLM 5.3 · Qwen 3.8 2.4T · DeepSeek V4 Pro · DeepSeek V4 Flash | EU-provider models are suitable for workflows where regional processing and transparent provider choice are central. |
| Document and image analysis | Claude Opus 5 · Gemini 3.7 Flash · Claude Opus 4.8 · GPT-5.6 Terra · Kimi K3 · GLM 5.3 Flash | Strong multimodal capabilities and large context windows for files, spreadsheets, screenshots, and complex documents. |
| Image generation | GPT-Image-2.0 · FLUX.2 [max] | Specialist models create visual drafts, marketing images, and creative variants directly from prompts. |
Start now and use EU-provider models and frontier models through one unified platform.