Claude

Claude is a family of large language models (LLMs) from Anthropic, a company with a strong focus on safety (Constitutional AI) and precise content fidelity.

---

Core concept

An LLM with a very large context window (up to 200k tokens), particularly adept at analysing long documents and generating high-quality code. Claude is considered less verbose and more precise than competing models.

Assessment

• Use case: Analysis of IT documentation, code generation, and creation of specialist content (such as this wiki).
• Advantage: Very high reliability in following complex instructions (system prompts) and excellent German-language capabilities.
• Limitation: API availability is still geographically restricted in some areas; focus is on text, code, and images; web search is available as a tool feature on supported models and platforms and must be enabled.

Related topics

• AI Development, the development context for AI-assisted work with Claude.
• GenAI and RAG, the pipeline in which Claude is applied.
• OpenAI, the service comparison point for Claude.
• Language Models, the overview of the model landscape.