Zed with mittwald
Zed is a high-performance code editor with built-in AI capabilities that supports OpenAI-compatible endpoints.
Prerequisites
- Zed editor installed
- mittwald AI Hosting API key (how to get one)
Configuration
Via UI
- Open Agent Panel settings (
agent: open settings) - Under "LLM Providers", click Add Provider
- Choose "OpenAI API Compatible"
- Configure:
- Provider name:
mittwald - API URL:
https://llm.aihosting.mittwald.de/v1 - API Key: your mittwald AI Hosting API key
- Model:
DeepSeek-V4.1-Flash-1017 - Context window:
1048576
- Provider name:
Via settings.json
Alternatively, configure Zed via settings.json:
{
"language_models": {
"openai_compatible": {
"mittwald": {
"api_url": "https://llm.aihosting.mittwald.de/v1",
"available_models": [
{
"name": "DeepSeek-V4.1-Flash-1017",
"display_name": "DeepSeek-V4.1-Flash-1017",
"max_tokens": 1048576,
"max_output_tokens": 32768,
"capabilities": {
"tools": true,
"images": true,
"parallel_tool_calls": false,
"prompt_cache_key": false
}
}
]
}
}
}
}
max_output_tokens leaves room for the reasoning that DeepSeek-V4.1-Flash-1017 writes before
it answers, and images lets you attach screenshots. The model is in the
Testphase. For production use, add Qwen3.8-27B-NVFP4 (Stable) with
"max_tokens": 256000 instead.
Usage
Once configured, you can use mittwald's AI models in Zed's Agent Panel for:
- Code generation and completion
- Refactoring and optimization
- Documentation and explanations
- Code review and suggestions
Next Steps
- Explore other IDEs - Compare different integrations
- Learn about MCP - Infrastructure management
- Best practices - Optimize AI usage