Centralized Req step that extracts token usage from provider responses, normalizes usage values across providers, computes costs, and emits telemetry.

This step:

• Extracts token usage numbers from provider responses
• Normalizes usage data across different provider formats
• Calculates costs using ReqLLM.Model cost metadata
• Stores usage data in response.private[:req_llm][:usage]
• Emits telemetry events for monitoring

Usage

request
|> ReqLLM.Step.Usage.attach(model)

Telemetry Events

Emits [:req_llm, :token_usage] events with:

• Measurements: %{tokens: %{input: 123, output: 456, reasoning: 64}, cost: 0.0123}
• Metadata: %{model: %ReqLLM.Model{}}