chat method
Single-turn chat (native Ollama API). Returns response text with any
<think>…</think> reasoning block stripped (user-visible content only).
For full access to the reasoning trace, use chatWithThinking.
Implementation
Future<String> chat(String model, String prompt, {String? system}) async {
final result = await chatWithThinking(model, prompt, system: system);
return result.content;
}