Token Cost Estimator

This tool estimates the character count, token count, and approximate USD cost for a given piece of text using standard pricing for a specified language model. It accepts two inputs: text (a required string containing the full content to analyze) and model (a required enum string selecting one of the supported models: gpt-4o, gpt-5-mini, gpt-5-pro, claude-3-5-sonnet). The worker uses these inputs to compute a structured token and cost estimation result. The final output must strictly validate against the system-provided structured output schema (auto-produced, not supplied here), with correct types and no extra keys.