The language model component inside an AI agent that receives text context and produces text responses, including tool calls.
↑ ContentsDefinition
The model receives context (system prompt, conversation history, tool results) and produces a response (text, tool calls, or both). It does not execute tools, read files, or run commands — it produces text that the runtime interprets. When the response contains tool calls, the runtime executes them. In an agent loop, the model is one component; the agent is the whole system: model, runtime, and tools. The model produces, the runtime executes.
↑ Contents