Rindler Review (2026): Pricing, Features & Honest Verdict

Rindler is a web automation service that signs into the portals your team uses and does the busywork there. You describe the job in plain English, it runs on sites with no API and hands back clean records. Charging per completed run is the honest model. $100 a month for one site is steep. Best for: ops teams drowning in portal work. Price: From $100/mo with a 7-day free trial (free plan: no). Rating: 7.5/10.

What is Rindler?

Rindler goes and does the web work your team still does by hand. Downloading statements and invoices, pulling records out of a supplier portal, checking order statuses, submitting forms, filing applications: the category of work that exists because some vendor built a login-gated website in 2011 and will never build an API for it. You describe the task the way you would explain it to a new hire, sign in to the site once, and Rindler runs it from then on, either on demand or on a schedule.

The company was founded in 2025 by Michael Serrano and Arthur De Los Santos, who met at MIT in 2022 and spent their undergrad years building things together, with research background at CSAIL. It is backed by Y Combinator. Worth noting for anyone reading the Product Hunt listing: Garry Tan appears on the maker list and is YC’s chief executive, not a founder of this company. The origin story on the about page is specific enough to be believable, which is that they kept meeting teams stuck on the same problem, a portal they log into daily, owned by someone else, with no API and no prospect of one, so people rekey the data by hand at volume in exactly the situation where a mistake is expensive.

What Are Rindler’s Key Features?

Pre-mapped portals instead of an agent rediscovering the page

This is the actual product and it is the thing that differentiates it from every browser agent shipped in the last eighteen months. Rindler walks a site before you ever use it, so a run follows a route it already knows rather than an LLM staring at a screenshot working out which button to click. Their words: there is no thinking time, the site is already understood, so the clock only runs on the work. If the site you need is not one they have been through, going through it is part of the plan rather than a quoted professional services project.

One finished task is one run, one dollar

The billing unit is a completed task, not a step, a token or a minute. Signing in, clicking through six screens, applying a filter and downloading nine statements is one run. The run stays open while the work goes on, so a long job never costs more than a short one, and if it comes back with nothing you are not charged. The rate is identical on every tier, so $1,000 buys ten times as many runs than $100 does, not cheaper ones. That last bit is unusually straight for SaaS pricing and it removes the standard game where the entry tier is deliberately bad value.

Permissions per action and per site

Every category of action is set to allow, ask first, or never, individually and per site. The categories are granular in the ways that matter: read pages and pull information, download a document, send a message or email, submit a form or application, upload or share a file, pay or check out or place an order, and change payment or billing details. Change one and it applies immediately. Given that this thing is signed into your bank and your suppliers, being able to set “read anything, never pay anything” per site is not a nice-to-have.

When the portal changes, they fix it centrally

Sites redesign, and that is where recorded click-path RPA dies. Rindler repairs the route once, for everyone using that site, and you do not file a ticket or pay for the repair. Their example screen shows a statements page that moved and was repaired at 2:14am while everything else stayed unchanged. This is the strongest argument in the whole product for buying instead of building, because a script your team wrote breaks on your side and becomes your problem on a Monday morning.

How Fast Is Rindler Really?

Rindler publishes a benchmark against browser-use, the open-source browser agent, and it is worth reading closely because it is more honest than most vendor benchmarks and still needs a caveat. The setup: same agent, same two grocery tasks, 38 runs, early 2026, and the only variable changed was whether the agent had a mapped site to work with. The results claimed are 4x faster task completion, 3x cheaper to run because it returns clean records instead of whole pages, and 2x fewer failed tasks.

The good part is that they named the comparison, the task count and the single variable, which is more disclosure than almost anyone in this category offers. The caveat is that it is a self-run benchmark on two tasks against one open-source baseline, published by the company selling the mapped sites. Two grocery tasks is not a portfolio, and the interesting question is how the mapped-route advantage holds up on a gnarly government or insurance portal rather than a consumer shopping site. Treat the direction as credible and the multipliers as unverified.

What Happens To Your Portal Logins?

You sign in yourself, on the real site, inside a managed browser, and complete any MFA. What Rindler keeps is encrypted, scoped to you and that one site, and per their FAQ never shown to the models. You can revoke it at any time. On some sites where the login is bound to a single live browser session, the credential is held per task rather than persisted. Every run keeps every step and every screen it saw, so a 3am run that could not finish comes back naming the portal and the step where it stopped.

The gap: Rindler is not SOC 2 certified, and says so plainly in its own FAQ rather than burying it. There is a security page with encryption, isolation and subprocessor detail, and they commit to answering security questionnaires within a week. For a two-year-old company that is a reasonable position honestly stated, and the plain “not yet” earns them more credit than a vague trust-page would. It is still a hard blocker if you are in a regulated industry, and it is a real risk to weigh given the credentials involved are for your bank, your ATS and your suppliers.

How Much Does Rindler Cost?

Starter is $100 a month with a 7-day free trial, covering one login-gated site set up with you, access to every already-supported site, 100 one-off runs a month and email support. Teams is $1,000 a month for 10 login-gated sites, 1,000 runs one-off or scheduled, saved automations you can put on a schedule, and a team workspace. Enterprise is custom, adds more sites public and login-gated, extra capacity above your included runs, dedicated onboarding, and runs on an annual agreement billed monthly with the first month as a pilot.

A run costs a dollar at every tier. That is easy to reason about and easy to compare against the alternative, which is a person doing it. If someone on your team spends three hours a week downloading statements from four portals, Starter pays for itself in the first fortnight. The problem is the entry point. One hundred dollars a month for a single site and 100 runs is a lot of commitment for a first purchase, and the jump to Teams is a 10x price step with no tier in between. A small ops team that needs three portals and 200 runs a month has no plan that fits them.

Compare that to the infrastructure layer, where Browserbase runs Free, then Developer at $20 a month, Startup at $99 and Scale custom, and Browser Use sells managed browsers at $0.02 an hour with tiers at $0, $29, $299 and $999. Those are much cheaper because they are selling you a browser and you are still writing and maintaining the automation. Rindler is selling the finished job with the site maintenance included, which is a different product at a different price. Whether the premium is worth it depends entirely on whether anyone on your team wants to own a fleet of scripts against portals that keep moving.

Who is Rindler Best For?

Use Rindler if your team logs into the same handful of portals every day and rekeys what it finds. Procurement and accounts payable pulling invoices and order statuses from supplier portals. Recruiters getting candidate data out of an ATS with no usable API. Anyone dealing with government or court filing systems, insurance portals or bank statement downloads. The common shape is high-volume, login-gated, no API, and expensive when someone typos a number. If that describes a real cost centre in your business, $100 a month is not the question, the question is how fast they can map your sites.

Skip Rindler if the site you care about has an API, because you should just use the API and it will be cheaper and more reliable. Skip it if you are a developer who wants to build your own agent, because Browserbase or Browser Use will cost you a tenth as much and give you the control you are after. Skip it if you need SOC 2 today. And skip it if your volume is genuinely low, because five runs a month at $100 works out at $20 a run and you would be better off doing it by hand.

Best Rindler Alternatives

Browserbase is the closest thing to a default in managed browser infrastructure. Free tier includes 3 concurrent browsers, 1 browser hour, 3 agent runs, 1,000 search calls and 1,000 fetch calls with 7 days of data retention. Developer is $20 a month, Startup is $99, Scale is custom. It is a building block rather than a service: you get the browsers and the session infrastructure, and the automation, the login handling and the repair work when a site changes are all yours.

Browser Use is the open-source agent Rindler benchmarks against, with over 80,000 GitHub stars, YC backing and a $17M seed led by Felicis. The cloud offering sells managed browsers at $0.02 an hour with no subscription required, and paid tiers at $29, $299 and $999 a month above a free tier. Cheapest way to get something working this week, and it is the option if you want the agent to figure out sites on the fly rather than paying someone to map them.

Skyvern sits between the two, offering a visual workflow builder, SOP uploads, browser recording, a copilot chat and native SDKs for browser automation. It is free to start with no credit card and 5,000 free credits every month. It publishes no paid ladder on its pricing page, so you will need to talk to them or sign up to find out what production costs, which is a strike against it if you are trying to budget.

Final Verdict: Is Rindler Worth It?

Rindler is betting that the interesting problem in browser automation is not the agent, it is the maintenance. That is a good bet. Every team that has tried to automate a supplier portal has the same story: it worked for four months, the vendor shipped a redesign, and now a script nobody owns is silently failing. Pre-mapping the sites and taking the repair burden centrally is a real product decision with real cost behind it, and the pay-per-completed-run pricing, with credits returned when a run comes back empty, is the fairest model in the category. The permission grid is the right shape for something holding your bank login.

What holds it back is the entry price and the youth of the company. One site and 100 runs for $100 a month is a big first cheque with a 10x gap to the next tier and nothing in between, which will lose them small teams who would otherwise be perfect customers. There is no SOC 2, the founding team is two people, and the benchmark supporting the core performance claim is two grocery tasks they ran themselves. If you have a portal problem that costs you real hours every week, take the 7-day trial and put your ugliest site in front of them, because that is the test that matters. If you are just curious about browser agents, start with Browser Use and come back when the manual work is actually hurting.