Pirsch vs Umami for AI agents

As of 12 Oct 2026, Umami gets an AI agent further: L3 · Does the work against L1 · Needs a person.

PirschL1Needs a person
  • Find: Some docs for agents: An MCP server.
  • Access: Can't sign up.
  • Use: Work not reached.
  • Pay: Payment not reached.
  • Leave: Leaving not reached.

Stopped at sign up: needs a human. Next: let an agent sign up on its own

Needs a person · L1 stops our agent cold. It read the public docs headlessly, then reached pirsch.io/signup and hit an ALTCHA "I'm not a robot" checkbox: our agent does not tick a bot check itself, and no sign-up API is documented, so verification, keys and the first call were never reached. The REST API behind that wall covers account work end to end, including deleting the account, and a third-party MCP server is linked from the docs. None of it was exercised: work done stayed at nothing. Paying runs through a hosted Stripe checkout, and the trial needs no card, but a person still has to get past the signup form first.

Driven 12 Oct 2026 · The drive and its evidence
UmamiL3Does the work
  • Find: Some docs for agents: llms.txt and an MCP server.
  • Access: Signs up in a browser.
  • Use: 75% of the work.
  • Pay: Payment not tried: robots.txt keeps our agent off the payment page for now.
  • Leave: Leaves on its own: Nothing to cancel, deletes its account.

Got in, in a browser; the API is paid. Next: do 80% of the core work (now 75%)

Gets in on its own: L3 · Does the work. Our agent signs up for Umami Cloud without a bot check and reaches the dashboard unaided. It stopped at Settings > API keys, which answers "API access requires a Pro plan." The documented send endpoint took hits with no key, but dropped them as bot traffic, leaving every report empty. In the browser it added a website, set goals and funnels, made shares and links, then deleted the account: login after that returned invalid-credentials, though a short link still redirected minutes later. Upgrading redirects to Stripe's checkout, which robots.txt keeps our agent off, so whether the trial needs a card is unknown.

Gets furtherDriven 8 Oct 2026 · The drive and its evidence

The two summaries: written and checked by our agents from each product's drive. How we write.

Where each one stops.

PirschStopped at sign up: needs a human

UmamiGot in, in a browser; the API is paid

Pirsch and Umami, checkpoint by checkpoint
CheckpointPirschUmami
DocsHeadlessHeadless
Sign upNeeds a humanIn a browser
VerifyNot reachedIn a browser
KeyNot reachedBlocked
First callNot reachedBlocked
First actionNot reachedIn a browser
PayNot reachedNot tried
CancelNot reachedNot needed
DeleteNot reachedIn a browser

Headless or in a browser.

How our agent passed each checkpoint, and what each product offers an agent.

Checkpoints by route, API, MCP, CLI and SDK
Checkpoints by how our agent passed them, and what each product offers an agent
PirschUmami
Headless (checkpoints)11
In a browser (checkpoints)04
Needs a human (checkpoints)10
Blocked (checkpoints)02
Not tested (checkpoints)71
Work doneNot measured75% of the core
APIYesYes
OpenAPINoNo
MCP serverYes (third-party)Yes (official)
CLINoNo
SDKYesYes
llms.txtNoYes

Bot checks, and where a person was needed.

Walls our agent met
PirschUmami
Bot checkALTCHA (proof of work) stopped it at sign up, with a visible challengeNone met
Wall typeA challenge on the formNo bot-check wall
First wallALTCHA checkbox on the sign-up formGot in, in a browser; the API is paid
A person helpedNoNo

A challenge on the form: It asks every visitor to prove they are human. It stops every agent on any network, and a signature doesn't help: a person does this step, or the owner opens it for agents. Where we drive from.

Level and part scores.

Pirsch and Umami: level and part scores, 0 to 100
PirschUmami
LevelL1 · Needs a personL3 · Does the work
Holds upNot measured64 of 100, fair
Agent-ready setup45 of 100, poor75 of 100, fair
StandardsMCP spec of 28 Jul 2026 (current) · MCP discovery (current)None recorded
SecurityNo review in this driveReported privately to the owner

Docs match is measured in an Audit. What the scores mean.

Pay, cancel and delete.

Paying, cancelling and deleting the account
PirschUmami
PayPayment not reachedPayment not tried
Free plan or trialTrial, no cardFree plan
CancelNot reachedNot needed
Delete the accountNot reachedIn a browser
How each one takes payment
  • Pirsch: Payment not reached.
  • Umami: Not tried yet: robots.txt keeps our agent off the payment page for now.

Pirsch vs Umami: the short answers.

Which can an AI agent use, Pirsch or Umami?

Umami gets an AI agent further: L3 · Does the work against L1 · Needs a person. With Pirsch our agent was stopped at sign up (needs a human); with Umami it got in, in a browser (the API is paid). Driven 12 Oct 2026 (Pirsch) and 8 Oct 2026 (Umami).

Can an AI agent sign up for Pirsch and Umami on its own?

Pirsch: no, needs a human. Umami: yes, in a browser.

Do Pirsch and Umami have an MCP server?

Pirsch: yes (third-party). Umami: yes (official).

Can an AI agent delete its Pirsch or Umami account?

Pirsch: not reached. Umami: yes, in a browser.

The drives and the category.

Put your own shortlist side by side.

A Stack check drives the tools you're choosing between and shows which ones your agents can use. $490 for three tools, ready in 24 hours.