Free
01
Start with journeys, simulations and readiness at no cost.
$0/ month
Includes
- Up to 100 agent journeys / month
- 10 simulation runs / month
- Weekly Agent Readiness scans
- Recommendations from all three
- 1 product surface
- Community support
Pay for journeys,not seats.
Start free with 100 agent journeys and 10 simulation runs every month, with Agent Readiness and Recommendations included. After that, pay only for what your agents actually do. No seat fees, no annual contract, no sales call to get started.
Free tier
100 journeys + 10 runs / month, forever
No card
Start without a credit card
Billing
Monthly, cancel anytime
Limits
Set a hard spend cap per project
Plans
Hit a limit?Add a card
Pay as you go
02
Pay only for the journeys and runs you use.
$0.10/ journey
Includes
- Unlimited journeys, billed per journey
- Simulation runs at $1 each after the free 10
- Agent Readiness and Recommendations included
- Docs, auth, APIs and MCP surfaces
- Priority support
Four products, one bill.
Journeys and simulation runs draw on one free monthly allowance per organization, then bill per use. Agent Readiness and Recommendations come with every plan.
| Product | Free every month | After that |
|---|---|---|
| /01Agent journeysReconstruct what real agents tried to do across your website, docs, APIs, SDKs and MCPs, and where they broke. | 100 journeys | $0.10/ journey |
| /02SimulationsEvalsRun real coding agents on your preference and experience prompts in clean sandboxes. See whether they pick you, and whether they finish the task. | 10 runs | $1/ run |
| /03Agent ReadinessScan your website, docs and app the way agents see them, with and without JavaScript. Every finding comes with its evidence and a fix prompt. | Weekly scans, plus Scan now | Included |
| /04RecommendationsRanked fixes built from your journeys, readiness scans and simulation runs, each with a prompt or pull request to ship it. | On every plan | Included |
Who we work with
From developer-tool startups to platform teams at scale, we work with companies whose products are increasingly used by AI agents, not just people.
Our pricing is designed tomake you happy
Most pricing pages are built to extract as much as possible. Ours is not. We charge for the journeys we reconstruct and the simulations we run, not seats, not platforms, not a sales process.
- 100 free agent journeys and 10 simulation runs every month, forever
- Ten cents a journey and a dollar a run after that, no annual contract
- Agent Readiness and Recommendations on every plan, never upsold
- No seat fees when someone new joins the workspace
- A hard spend cap when you want one
- Cancel anytime and keep the free tier
Billing questions, answered.
What counts as an agent journey?
One continuous session where an AI agent tries to complete a task on your product, from first request to success or abandonment. Retries inside the same session are part of the same journey.
What counts as a simulation run?
One agent working through one of your preference or experience prompts, from start to finish, in a clean sandbox. Three samples of the same prompt, or the same prompt on three agents, are three runs.
Do Agent Readiness and Recommendations cost extra?
No. Every workspace gets an automatic readiness scan each week, and anyone can press Scan now in between. Recommendations are built from the journeys, scans and runs you already have, so there is nothing separate to meter.
Do free allowances roll over?
No. Your free allowance resets on the 1st of each month. Unused free journeys and simulation runs don’t carry forward.
Can I cap my spend?
Yes. Set a monthly cap per project. When you reach it, Agentrail keeps collecting data but pauses billable reconstruction and simulation runs until the next cycle or until you raise the cap.
How do volume tiers work?
Tiers are graduated. Each tier’s price applies only to the units inside it, so your first 10,000 journeys are always billed at the lower tiers before higher volume kicks in.
Do you offer discounts for startups or open source?
Yes. Early-stage startups and open-source maintainers can apply for credits. Talk to us and tell us what you are building.
