httpcdpagent-toolsthirdweb

fetchharbor.benlab.download

Generate one bounded assistant response with the operator's self-hosted Ollama model. Use for short, single-message inference where local processing is preferred. Accepts up to 8,000 characters and returns the model name, response text, and token counts when Ollama supplies them.