Skip to main content

Understanding Usage Costs

This guide explains how usage costs work in Hunch, including AI token consumption and model pricing.

How Billing Works​

Hunch usage is based on:

  • Sessions: Number of chat conversations
  • AI Tokens: Input and output tokens used by AI models
  • Additional features: Extra services and integrations

Session Usage​

Each visitor interaction counts as one session. A session includes:

  • All messages between visitor and AI
  • Context maintained during the conversation
  • Any human handoff that occurs

AI Token Costs​

AI models process text in "tokens" - roughly 1 token = 1 word. Different models have different pricing:

GPT-4o (OpenAI)​

Token TypePrice per 1M tokens
Input$2.50
Output$10.00

GPT-4o-mini (OpenAI)​

Token TypePrice per 1M tokens
Input$0.15
Output$0.60

Claude 3.5 Sonnet (Anthropic)​

Token TypePrice per 1M tokens
Input$3.00
Output$15.00

Claude 3 Haiku (Anthropic)​

Token TypePrice per 1M tokens
Input$0.25
Output$1.25

Gemini 1.5 Pro (Google)​

Token TypePrice per 1M tokens
Input$1.25
Output$5.00

Gemini 1.5 Flash (Google)​

Token TypePrice per 1M tokens
Input$0.075
Output$0.30

Estimating Costs​

Example calculation for 1,000 sessions with average usage:

  • Average input per session: 500 tokens
  • Average output per session: 300 tokens
  • Total input: 500,000 tokens
  • Total output: 300,000 tokens

Using GPT-4o-mini:

  • Input cost: (500,000 / 1,000,000) × $0.15 = $0.075
  • Output cost: (300,000 / 1,000,000) × $0.60 = $0.18
  • Total per 1,000 sessions: ~$0.26

Using GPT-4o:

  • Input cost: (500,000 / 1,000,000) × $2.50 = $1.25
  • Output cost: (300,000 / 1,000,000) × $10.00 = $3.00
  • Total per 1,000 sessions: ~$4.25

Model Selection​

Choose the right model for your needs:

  • GPT-4o-mini / Gemini 1.5 Flash: Cost-effective for simple queries
  • GPT-4o / Claude 3.5 Sonnet: Best for complex conversations
  • Claude 3 Haiku: Balanced option for moderate complexity

Configure your default model in Settings > AI Settings.

Usage Dashboard​

Monitor your usage in the dashboard:

  • Current month's token consumption
  • Session count vs plan limit
  • Cost breakdown by model
  • Daily/weekly trends

Staying Within Limits​

To avoid overages:

  1. Monitor your usage regularly
  2. Set up usage alerts in billing settings
  3. Consider upgrading your plan for higher limits
  4. Use cost-effective models for simple queries

Billing Cycle​

Usage is billed monthly:

  1. You start each month with your plan's included usage
  2. Overage charges apply if you exceed limits
  3. Invoice generated at end of billing cycle
  4. Payment processed automatically

View your detailed usage and billing in Settings > Billing.