How much does it cost to run an AI chatbot?

It depends on volume and model. A chatbot handling 1,000 messages/day with ~500 input and ~300 output tokens per message costs $0.90-$45/month on budget models like Gemini 2.0 Flash Lite or GPT-4o mini, but $75-$450/month on premium models like GPT-5.5 or Claude Opus 4.7. Use our calculator above to model your exact usage.

What is the cheapest AI model for a chatbot?

Google Gemini 2.0 Flash Lite ($0.075/$0.30 per 1M tokens) and Llama 3.1 8B ($0.10/$0.10) are the cheapest options. For better quality at a low cost, GPT-4o mini ($0.15/$0.60) and DeepSeek V4 Flash ($0.14/$0.28) offer excellent value. The right choice depends on your quality requirements.

How do I reduce my chatbot API costs?

Key strategies: (1) Use cheaper models for simple queries and reserve premium models for complex ones (model routing), (2) Cache frequent questions to avoid duplicate API calls, (3) Limit output tokens to prevent runaway generation, (4) Use prompt templates to reduce input token count, (5) Consider batch API endpoints where providers offer 50% discounts.

Which is cheaper for chatbots: ChatGPT or Claude?

At the budget tier, GPT-4o mini ($0.15/$0.60) is cheaper than Claude Haiku 4.5 ($1.00/$5.00). At mid-tier, GPT-5 ($1.25/$10.00) undercuts Claude Sonnet 4 ($3.00/$15.00). However, DeepSeek V4 Flash ($0.14/$0.28) beats both budget options. Use our calculator to compare with your specific token counts.

How many tokens does a typical chatbot message use?

A typical customer support chatbot uses 300-800 input tokens (system prompt + conversation history + user message) and 100-400 output tokens (assistant response). A sales chatbot with longer responses uses 500-1500 output tokens. An internal assistant with RAG context uses 1000-3000 input tokens.

AI Chatbot Cost Calculator

Building a chatbot? See exactly what it'll cost across 33 AI models. Enter your expected messages and token counts — get instant monthly cost estimates and find the cheapest provider for your volume.

Chatbot type:

Volume:

System prompt + user message (tokens)

System prompt + conversation history + user message

Assistant response (tokens)

Typical chatbot response: 100-500 tokens

Messages per day

Days per month

Your Chatbot Cost

Cost per message $0.0000

Cost per 1K messages $0.00

Input cost (monthly) $0.00

Output cost (monthly) $0.00

Monthly Total $0.00

All 33 Models Ranked by Chatbot Cost

Sorted from cheapest to most expensive for your current settings. The green badge shows how much you save vs. the most expensive model.

#	Model	Provider	Tier	Cost/Message	Monthly Cost	Annual Cost	Savings vs Most Expensive

💡 Key Insight

Want to optimize your chatbot costs?

Pro gives you model routing recommendations (use cheap models for simple queries, premium for complex ones) that can cut your chatbot API bill by up to 40%.

$29

one-time, lifetime

40%

avg. savings

Unlock Pro Features — $29

14-day money-back guarantee. Lifetime access.

How to Estimate Your AI Chatbot Costs

Running an AI chatbot means paying per API call. Every message costs tokens — the system prompt, conversation history, user message, and assistant response all count. Here's how to plan your budget:

Typical Token Counts by Chatbot Type

Customer support bot: 300-800 input tokens (system prompt + history + question), 100-400 output tokens (answer)
Sales/lead gen bot: 500-1,200 input tokens, 200-800 output tokens (longer, persuasive responses)
Internal assistant (RAG): 1,000-3,000 input tokens (context documents), 200-600 output tokens
Creative writer: 200-600 input tokens, 500-2,000 output tokens (long-form generation)

Model Selection Strategy

Not every message needs a premium model. The most cost-effective approach is model routing:

Tier 1 (simple queries): Route to budget models like GPT-4o mini ($0.15/$0.60) or Gemini 2.0 Flash ($0.10/$0.40) — handles ~60% of chatbot traffic at 1/10th the cost
Tier 2 (complex queries): Use mid-tier models like GPT-5 ($1.25/$10.00) or Claude Sonnet 4 ($3.00/$15.00) for ~30% of traffic
Tier 3 (critical queries): Reserve premium models like GPT-5.5 ($5.00/$30.00) or Claude Opus 4.7 ($5.00/$25.00) for ~10% of traffic

This hybrid approach typically costs 40-60% less than using a single model for everything.

Cost-Saving Tips for Chatbot Builders

Cache frequent questions: If 30% of users ask the same 5 questions, cache those responses and save 30% on API costs
Optimize system prompts: A 500-token system prompt on every message adds up fast — keep it concise
Set max output tokens: Prevent runaway responses by setting a token limit (e.g., 500 tokens max)
Use conversation summarization: Instead of sending full history, summarize older messages to reduce input tokens
Batch API endpoints: OpenAI offers 50% discounts for batch processing — ideal for non-real-time chatbot tasks

Want to compare two models side by side?

Use the Comparison Tool →

Related Tools

AI API Cost Calculator — General-purpose calculator with all 33 models
Model Comparison — Side-by-side model comparison
Cost Explorer — See all 33 models ranked by cost
Cost Optimizer — Get personalized savings recommendations
Model Switch Calculator — See savings from switching models
Budget Planner — Step-by-step wizard for your AI budget
Model Selector Quiz — Answer 5 questions to find the right model