| Capability | -Description | -
|---|---|
vision |
- Process and understand images | -
functionCalling |
- Use tools and execute actions | -
jsonOutput |
- Generate valid JSON responses | -
family |
- Model architecture type (e.g., "gpt", "llama") | -
structuredOutput |
- Generate formatted responses | -
multipleSystemMessages |
- Support multiple system prompts | -
| Provider | -ModelInfo | -Note | -
|---|---|---|
| OpenAI | -Optional | -Pre-configured for standard models | -
| Anthropic | -Optional | -Pre-configured for Claude models | -
| Ollama | -Required | -Must specify all capabilities | -
| Custom | -Required | -Must specify all capabilities | -