AI Usage

Overview

The AI Usage section gives you a full picture of all AI activity happening inside your bot. All data is displayed based on your workspace timezone and can be filtered by a custom date range. Each metric is compared to the previous period so you can track changes over time.

Status

This section shows the total number of successful and failed AI calls within the selected date range.

Bar Chart:

• The teal bars represent successful AI calls.

• The orange bars represent failed AI calls.

A donut chart on the right shows the percentage split between success and error.

Open

Requests by Provider

This section shows the total number of AI requests broken down by provider over the selected date range. The available providers are OpenAI, DeepSeek, XAI, Claude, Gemini, Groq, and Ainvented. 

A donut chart on the right shows the percentage share of each provider.

Open

Average Process Time by Provider (ms)

This chart shows the average response time for each provider, measured in milliseconds, plotted over the selected date range. It helps identify which providers are responding faster or slower during a given period.

Open

Total Tokens

Shows the total number of tokens consumed across all AI calls. Includes a line chart comparing the current period to the previous period.

Open

Average Prompt Tokens per Request

Shows the average number of input tokens sent per AI request. Includes a comparison between the current and previous period.

Average Completion Tokens per Request

Shows the average number of output tokens returned per AI request. Includes a comparison between the current and previous period.

Open

Average Process Time (ms)

Shows the average time taken to complete an AI request across all providers. Includes a comparison between the current and previous period.

Process Time Breakdown

A table showing what percentage of requests fell within each processing time range. The time ranges are: under 1s, 1s to 2s, 2s to 3s, 3s to 4s, 4s to 5s, 5s to 10s, and over 10s.

Open