Model metadata & transparency
Steinkauz shows metadata for each model so you can see where it comes from, what it can do, and how it is priced. This supports transparent, informed choices.
System-provided metadata
Steinkauz attaches metadata to each model so you can compare and choose without guessing. This data is maintained by the platform and is the same for all users.
- Lab — Who created or trained the model (e.g. Anthropic, OpenAI, Google). Shown with the lab name and, when set, a country-of-origin flag.
- Provider — Who serves the model for inference (the API backend you route through). Shown with the provider name and, when set, a country-of-origin flag. Lab and provider can differ — for example, a model created by one lab may be served via another provider or the Gateway.
- Modalities — What the model can accept as input and produce as output: Text, Image, Audio, Video, PDF. For example, an image-generation model will list Image in its output modalities; a chat model is typically text in and text out. Shown in the model details (e.g. the info popover next to the model name).
- Capabilities — Boolean flags that describe what the model supports: Reasoning (extended reasoning, e.g. chain-of-thought), Tool calling (can use tools such as web search or image generation), Structured output (can return JSON or other structured data). These help you pick a model that fits the kind of task you’re doing.
You can use this metadata in the model list and in Settings → Models to filter, sort, and compare models before starting a conversation.
User-provided metadata
You can personalize how you see and organize models:
- Tags — Add your own tags to models (e.g. “work”, “creative”) to find them quickly.
- Favorite — Mark models as favorites so they appear in a dedicated list or at the top of the selector.
- Enable/disable — Turn a model off to hide it from the model selector in chat; it remains visible in Settings → Models so you can turn it back on anytime.
Tags and favorites are stored per user and only affect your view within the active context.
Transparency and pricing
Model metadata is part of Steinkauz’s transparency approach: you see which lab created the model, which provider serves it, and which model you are using, and you can check usage, costs, and activity for your active context. Together with Data routing policy and provider configuration, metadata helps you control where your data goes and what you pay.