> ## Documentation Index > Fetch the complete documentation index at: https://docs.toolken.ai/llms.txt > Use this file to discover all available pages before exploring further. # Overview > See and slice your LLM metrics by agent, feature, or customer. Toolken records every request that goes through the gateway. You see cost, token usage, latency, and error rates in one dashboard. Tag a request and you can break any of those down by agent, customer, environment, or any property you choose. ## What you can measure What each request, agent, or customer spent. Input, output, and cached tokens. Latency and time to first token. Error rates and status codes. ## How it works You tag requests with metadata headers. Toolken turns each metadata key into a dimension you can group and filter by, across every metric above. ```bash theme={null} -H "X-Toolken-Metadata-Agent: research-agent" ``` A request with no tags still shows up by model and provider. Tags add any breakdown you need on top. ## Next Tag requests so you can slice by them. Group and filter any view. The views you get out of the box.