---
title: "Conversations summary"
method: GET
path: "/projects/{project_id}/agent/analytics/conversations_summary"
tags: ["Agent"]
---

# Conversations summary

`GET /projects/{project_id}/agent/analytics/conversations_summary`

## Request

- Base URL: `https://api.beyondwords.io/v1`
- URL: `https://api.beyondwords.io/v1/projects/{project_id}/agent/analytics/conversations_summary`
- Auth: API key in header `X-Api-Key`

Retrieves conversation summary stats. For conversational_agent (the default), this is ElevenLabs conversation count, average duration and average messages. For ai_assistant, this is distinct attributed conversation_id counts and tool-call attribution coverage.

## Query parameters

- `agent_type` 'conversational_agent' | 'ai_assistant'
- `time_range[start]` string
- `time_range[end]` string
- `access_tiers[]` string[]

## Response `200`

OK

- AgentAnalyticsConversationsSummaryModel
  - `conversations` integer: For conversational_agent, total ElevenLabs conversations in the requested time range. For ai_assistant, distinct attributed conversation_id values with connector tool calls in the range.
  - `attributed_tool_calls` integer: ai_assistant only. Number of tool calls carrying a usable conversation_id.
  - `unattributed_tool_calls` integer: ai_assistant only. Number of tool calls without a usable conversation_id, so they could not be grouped into a conversation.
  - `avg_duration_ms` integer, nullable: conversational_agent only. Average conversation duration in milliseconds. Null when conversations is 0, or when agent_type is ai_assistant (no per-conversation duration is recorded).
  - `avg_messages` number, nullable: conversational_agent only. Average number of transcript turns per conversation. Null when conversations is 0, or when agent_type is ai_assistant (tool-call counts do not map to chat messages).
  - `evaluated_conversations` integer: conversational_agent only. Number of conversations with a known (success or failure) outcome, excluding unknown outcomes. Denominator for success_rate. Always 0 for ai_assistant.
  - `success_rate` number, nullable: conversational_agent only. Share of evaluated conversations ElevenLabs marked as successful, between 0 and 1. Excludes conversations with an unknown outcome. Null when evaluated_conversations is 0, or when agent_type is ai_assistant (no outcome is recorded).
  - `series` object[]: Dense daily values, present only when aggregation=day was requested. The same conversation_id can appear on several days; do not sum per-day conversations to reconstruct the whole-range total.
    - `start` string, date-time: The start of the reporting day (ISO 8601, UTC)
    - `conversations` integer
    - `attributed_tool_calls` integer
    - `unattributed_tool_calls` integer
    - `avg_duration_ms` integer, nullable
    - `avg_messages` number, nullable
    - `evaluated_conversations` integer
    - `success_rate` number, nullable

## Other responses

- `404`: Agent beta not enabled for this organization
- `422`: missing time_range

## Changes

- **2026-11-03** `994b9853e685`: 1 info
  - endpoint added

[Operation history](https://skmtc.dev/beyondwords/apis/beyondwords-rest-api/changes/projects/:project_id/agent/analytics/conversations_summary/get.md)

---

[API](https://skmtc.dev/beyondwords/apis/beyondwords-rest-api.md) - [All operations](https://skmtc.dev/beyondwords/apis/beyondwords-rest-api/llms.txt) - [OpenAPI document](https://skmtc.dev/beyondwords/apis/beyondwords-rest-api/revisions/994b9853e685?raw)