Qwen3.8-27B: versatile AI for assistance, coding and agents on Swiss infrastructure

The Model at a glance:

• Model: Qwen3.8-27B, Open Weight, context length 256k tokens
• Applications: Assistance systems, automation and analysis, coding, document processing, agent-based workflows
• Coding: Code generation, analysis and review in over 20 programming languages, function calling, tool integration
• Integration: OpenAI-compatible API, directly into existing applications

Anyone wishing to use AI productively is often faced with a choice of models, each with its own strengths. Qwen3.8-27B is designed for the opposite scenario: a model that performs well across several areas and is suitable as a central foundation. stepping stone strongly recommends it as a well-balanced foundation for assistance, coding, multimodal tasks and agents.

Its versatility is evident in its typical areas of application. Teams use the model for chatbots and assistance systems, automation and analysis, internal knowledge search, multilingual analysis workflows, document processing and agent-based workflows. The large context size of 256k tokens allows extensive documents and longer conversation histories to be processed in a single run.

The model has a clear strength in coding. It generates, analyses and reviews code in over 20 programming languages. For agent-based setups, it supports function calling, tool integration and multi-step task planning. This makes it suitable both as a coding assistant in developers’ day-to-day work and as a foundation when AI agents need to call their own tools and plan multi-stage tasks themselves.

At the same time, Qwen3.8-27B remains entirely within Switzerland. The model runs on Swiss infrastructure in stepping stone’s data centres, and no data leaves the country. Companies remain independent of OpenAI, Google and Amazon and retain control over their AI strategy.

Integration is straightforward: via an OpenAI-compatible API, the model can be integrated directly into existing applications and workflows.

→ Find out more about Qwen3.8-27B