Skip to main content

Get in Touch

Have a suggestion, spotted a bug, noticed an incorrect algorithm, or just want to talk cubing? Every message reaches me directly. Feedback shapes what gets built next.

trevor@aicubetrainer.com

What to include

  • Bug reports: what you did, what happened, what you expected. Screenshots help.
  • Algorithm corrections: the case, the alg you see on the site, and the alg you'd recommend.
  • Feature requests: describe the training problem it solves for you.
  • Success stories: if the site helped you break a time barrier, tell me, those make hard days easier.

Replies usually come within a few days. This is an independent project run by one person, thanks for your patience.

What Happens When You Email

Every email goes directly to Trevor, the developer and sole person behind AI Speed Cube Trainer. There's no support ticket system, no outsourced help desk, and no automated replies. When you write, a real person reads it, usually within a day or two.

Bug reports almost always lead to a fix within the next deployment. Feature requests get added to a running list that directly influences what gets built next. Algorithm corrections are verified against established speedcubing sources and updated quickly. If you spotted something wrong, thank you, algorithm accuracy is the one thing this tool lives or dies by.

Common Questions Before You Write

Is there a way to reset my score or F2L progress?

Your data is stored in your browser's local storage. To reset it, go to your browser settings, find "Site data" or "Cookies and site data," locate aicubetrainer.com, and clear it. This resets everything including scores and spaced-repetition progress. A dedicated in-app reset button is planned for a future update.

The OLL detector returned "no match": is it broken?

It's almost always a mapping error, not a bug. The most common mistake is forgetting the back face side strip, or accidentally including stickers that aren't in the top row. The side strips represent only the top row of each of the four surrounding faces, not the full face. Re-read the cube carefully and try again. If you've triple-checked and still get no match, email with a screenshot.

Can you add support for [X method / X cube type]?

The trainer is currently focused on CFOP for the 3x3, specifically OLL, PLL, and F2L. Other methods (Roux, ZZ, Petrus) and cube types (4x4, megaminx, pyraminx) are possible future additions but are not currently planned. Send a feature request and it goes on the list. Popular requests get prioritized.

Is there an iOS or Android app?

No native app exists. The web trainer is built to be fully mobile-responsive and runs in any mobile browser. For the closest app-like experience, open the site in Chrome or Safari and use "Add to Home Screen", this creates a standalone launcher with no browser chrome.

Why is this free? What's the catch?

There's no catch. The site is free and is designed to be supported by display advertising once ads are approved and live, so hosting stays covered without a paywall. There are no premium tiers, no paid features, and no selling of personal data. If you find the site valuable, sharing it with another cuber is the most helpful thing you can do today.

I found an algorithm that's better than the one shown: how do I report it?

Email with: the case name (e.g., OLL 32 / T-perm / F2L case #14), the algorithm currently shown, and the algorithm you'd recommend with a source if you have one. Algorithms are chosen based on a combination of move count, execution ergonomics, and community consensus. If a genuinely better option exists, it gets added, often as an alternate rather than a replacement, since algorithm preference is personal.

A Note on Response Times

AI Speed Cube Trainer is an independent project maintained by one person alongside other work. Most emails get a response within 2–5 days. During active development periods, responses may be faster. During major competitions or busy periods, they may take a bit longer. If you haven't heard back in a week, a polite follow-up is welcome.

Messages that report a clear bug with reproduction steps get moved to the top of the queue. If something is broken, the goal is always to fix it as quickly as possible, especially anything affecting detection accuracy.