Token spend per agent run
Multi-step agents re-send context on every step, so input tokens pile up fast.
Home / Dev & AI / AI & LLM Cost / AI Agent Cost Calculator
AI & LLM Cost
AI Agent Cost Calculator
Autonomous agents call the model many times per task, so costs add up fast. Enter tasks per month, steps per task and tokens per step, pick a model, and see the true cost per task and per month.
Estimated monthly cost
$0
Cost per task—
Cost per 1,000 tasks—
LLM calls per month—
Input token cost—
Output token cost—
Annual estimate—
Your agent priced across AI models
Same task volume and step count, cheapest model first.
What the AI Agent Cost Calculator does
An AI agent does not answer in one shot. It plans, calls tools, reads results and reasons again, often six to twenty model calls for a single task. Each of those steps resends the growing context, so an agent can burn ten to fifty times the tokens of a plain chat reply. That is exactly why an agent cost calculator matters before you ship one.
This tool multiplies tasks by steps to get total model calls, prices the tokens each step moves, and shows the cost per task so you can see whether an agent is economical at your volume. It also compares models, since a cheaper model on a well designed agent can beat a premium model on a wasteful one.
Heads up on pricing. The rates built into this tool are public list estimates for July 2026 and can change without notice. Providers also offer batch discounts, prompt caching and volume tiers that lower real costs, so treat the output as a planning estimate, not a quote.
The formula
model calls = tasks × steps per task
cost = (calls × input tokens ÷ 1M × input rate) + (calls × output tokens ÷ 1M × output rate)
cost = (calls × input tokens ÷ 1M × input rate) + (calls × output tokens ÷ 1M × output rate)
Worked example
An agent handles 20,000 tasks a month at 6 steps each, so 120,000 model calls. At 2,500 input and 400 output tokens per step on GPT-5, input is 300M × $1.25 + output 48M × $10 per million = $375 + $480 = about $855 a month, roughly $0.043 per task.
AI model pricing used in this calculator (July 2026 estimates)
| Model | Provider | Input / 1M tokens | Output / 1M tokens |
|---|---|---|---|
| GPT-5 | OpenAI | $1.25 | $10.00 |
| GPT-5 mini | OpenAI | $0.25 | $2.00 |
| Claude Opus 4.1 | Anthropic | $15.00 | $75.00 |
| Claude Sonnet 4.5 | Anthropic | $3.00 | $15.00 |
| Claude Haiku 4.5 | Anthropic | $1.00 | $5.00 |
| Gemini 2.5 Pro | $1.25 | $10.00 | |
| Gemini 2.5 Flash | $0.30 | $2.50 | |
| DeepSeek V3 | DeepSeek | $0.27 | $1.10 |
| Grok 4 | xAI | $3.00 | $15.00 |
| Llama 4 Maverick | Meta | $0.20 | $0.60 |
Prices are public list estimates for planning as of July 2026 and change often. Providers bill per token, where roughly 1,000 tokens equals about 750 words. Always confirm live rates on the provider pricing page before you set a budget.
Ways to lower your AI costs
- Match the model to the task. Route simple requests to a small model like GPT-5 mini, Gemini Flash or DeepSeek V3, and keep premium models for complex reasoning where quality pays off.
- Cache repeated context. Prompt caching means you stop paying full price for the same system prompt or documents on every call.
- Cap output length. Output tokens cost several times more than input, so limiting response length is often the fastest saving you can make.
- Trim prompts and history. Summarize old conversation turns and drop context the model does not actually need to answer well.
- Batch non-urgent work. Many providers offer batch pricing at around half the standard rate for jobs that can wait a few hours.
Frequently asked questions
Why do AI agents cost more than chatbots?
Because one task triggers many model calls, and each call resends the accumulated context and tool output. Steps multiply tokens, so agents are far more token hungry than single turn chat.
How many steps does a typical agent take?
Simple agents finish in 3 to 6 steps, complex research or coding agents can run 10 to 30. Fewer, better planned steps usually beat many shallow ones.
How do I reduce agent costs?
Cap the step count, summarize context between steps, cache the system prompt and tool schema, and use a smaller model for routine steps while reserving a strong model for planning.
Do tool calls cost tokens?
The tool definitions and their returned results count as tokens on the next model call, so verbose tool output directly raises cost.
Are these numbers reliable?
They are July 2026 estimates for planning. Agent token use is variable, so measure real runs before you lock a budget.