> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.vapi.ai/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.vapi.ai/_mcp/server.

## New: Call Metrics & Artifact Improvements

You can now access detailed call performance metrics and structured output IDs directly from your call artifacts.

#### Turn Latencies

**Each conversation turn's latency**\
`Call.artifact.performanceMetrics.turnLatencies`

#### Model Latency (avg)

**Average time for the model to generate a response**\
`Call.artifact.performanceMetrics.modelLatencyAverage`

#### Voice Latency (avg)

**Average time to synthesize voice**\
`Call.artifact.performanceMetrics.voiceLatencyAverage`

#### Transcriber Latency (avg)

**Average time to transcribe voice**\
`Call.artifact.performanceMetrics.transcriberLatencyAverage`

#### Endpointing Latency (avg)

**Time to detect end of a conversation turn**\
`Call.artifact.performanceMetrics.endpointingLatencyAverage`

#### Turn Latency (avg)

**Average latency to complete a conversation turn**\
`Call.artifact.performanceMetrics.turnLatencyAverage`

#### Structured Output IDs & Results

**Track and extract structured outputs from your calls**

* **During call:** Access array of output IDs\

  `Call.artifactPlan.structuredOutputIds`
* **After call:** Extracted outputs are stored here\

  `Call.artifact.structuredOutputs`

> **Info**
>
> These improvements help you monitor, debug, and analyze your calls with greater detail.