Stay organized with collections
Save and categorize content based on your preferences.
This page contains current API pricing for the use of Insights.
This page will be updated to reflect any changes to these prices. We reserve the
right to change these prices.
Insights is priced according to the number of
conversations processed by the API. A conversation is text between multiple
participants. A conversation turn is when the conversation switches between
participants. For pricing purposes, a single conversation is defined as having a
maximum of 120 turns, with each turn having a maximum of 1000 characters. More
than 1000 characters in a single turn will cause that turn to count for multiple
turns. More than 120 turns in a conversation will cause the conversation to be
billed as more than one conversation. For example, if a conversation has 241
turns, it will be billed as 3 conversations.
Building a topic model
does not require any data to be labeled and is free. You need a minimum number of conversations imported into Insights to train a model. Importing audio data that needs to be transcribed by Speech-to-Text to get the transcript incurs a charge. After you have a model, you can use it to analyze conversations and label the topics, incurring topic-model labeling charges listed in the following table.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Hard to understand","hardToUnderstand","thumb-down"],["Incorrect information or sample code","incorrectInformationOrSampleCode","thumb-down"],["Missing the information/samples I need","missingTheInformationSamplesINeed","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2025-08-29 UTC."],[[["\u003cp\u003eInsights API pricing is based on the number of conversations processed, with a conversation defined as up to 120 turns, each with a maximum of 1000 characters.\u003c/p\u003e\n"],["\u003cp\u003eExceeding the conversation limits (120 turns) or turn limits (1000 characters) will result in additional charges.\u003c/p\u003e\n"],["\u003cp\u003eBasic Insights analysis, including sentiment, entities, and highlighters, costs $2 per 100 conversations.\u003c/p\u003e\n"],["\u003cp\u003eTopic model labeling is also priced at $2 per 100 conversations.\u003c/p\u003e\n"],["\u003cp\u003eQuality AI features for virtual/human agents cost $0.005 per conversation per score, and agent subscriptions are $50 per agent per month.\u003c/p\u003e\n"]]],[],null,["# Pricing\n\nThis page contains current API pricing for the use of Insights.\nThis page will be updated to reflect any changes to these prices. We reserve the\nright to change these prices.\n\nInsights is priced according to the number of\n**conversations** processed by the API. A conversation is text between multiple\nparticipants. A conversation **turn** is when the conversation switches between\nparticipants. For pricing purposes, a single conversation is defined as having a\nmaximum of 120 turns, with each turn having a maximum of 1000 characters. More\nthan 1000 characters in a single turn will cause that turn to count for multiple\nturns. More than 120 turns in a conversation will cause the conversation to be\nbilled as more than one conversation. For example, if a conversation has 241\nturns, it will be billed as 3 conversations.\n\nBuilding a [topic model](/contact-center/insights/docs/topic-modeling-overview)\ndoes not require any data to be labeled and is free. You need a minimum number of [conversations imported into Insights](/contact-center/insights/docs/topic-modeling#import_conversation_data) to train a model. Importing audio data that needs to be transcribed by Speech-to-Text to get the transcript incurs a charge. After you have a model, you can use it to analyze conversations and label the topics, incurring topic-model labeling charges listed in the following table."]]
