Agent Workflow Cost Calculator
Estimate the cost of a multi-step agent workflow.
With the default inputs, $0.0270 per workflow run — 6 steps · $810.00/month. Enter your own numbers below to recompute instantly; the full step-by-step math is shown under the worked example.
Results update automatically as you type.
- Steps
- 6
- Cost / step
- $0.00450
- Cost / run
- $0.0270
- Cost / month
- $810.00
Model a multi-step agent or pipeline: set the number of steps and per-step tokens to see the cost per run and per month.
How this is calculated
Cost per run = steps × (input per step × input price + output per step × output price), scaled by runs per month. It models a fixed multi-step pipeline where each step is its own model call with its own tokens, so cost scales linearly with the number of steps.
Is this a good result? What to do next
A workflow's cost is predictable and linear in steps — unlike an open-ended agent, you control exactly how many calls run. If a run is pricier than expected, it's usually carried-over context inflating per-step input, or retries.
Typical planning ranges
- Cost scales
- linearly with steps
- Each step
- its own call and tokens
- Surprises
- usually carried context or retries
Ranges are typical planning figures to sanity-check your result, not authoritative benchmarks. Your numbers will vary with use case, volume, and vendor.
How to improve this number
- Remove or merge unnecessary steps.
- Trim per-step input; don't carry context you don't need.
- Use a cheaper model for mechanical steps.
Common mistakes
- Carrying full context into steps that don't need it.
- Ignoring retries that add hidden step runs.
When to use a different approach
For an open-ended agent with tool calls, use the AI agent cost calculator. For carried-memory growth, use the agent memory cost calculator.
Worked example (defaults)
With the default inputs above, here is the result:
- Steps
- 6
- Cost / step
- $0.00450
- Cost / run
- $0.0270
- Cost / month
- $810.00
Sources & references
Frequently asked questions
How is a workflow different from a single call?+
Each step is its own model call with its own tokens, so cost scales with the number of steps.
Why is my agent more expensive than expected?+
Carried-over context and retries inflate per-step input tokens. See our agent memory calculator.
Related calculators
Answering a real question?
This calculator powers these problem-solving guides: