Photomath Review

Owned by Google since 2023, with 100M+ downloads on Google Play alone. This isn't a startup bet — it's an established consumer product with real distribution. Vendor viability isn't the concern here.

The multi-method solving feature is what separates it from a plain answer engine. Showing students elimination method, Cramer's rule, and comparison method side-by-side teaches reasoning, not just answers. That's the actual learning argument. Support for 32 languages makes it defensible for district-level or global deployments, though every scan needs an internet connection.

The tradeoff: free tier covers a lot, which means convincing students to pay $69.99/year is a real conversion problem if you're licensing institutionally. The pricing page doesn't show volume or institutional tiers, so bulk deployment math gets murky fast.

The multi-method problem solving feature is the real asset here. Showing elimination, Cramer's rule, and comparison side-by-side is legitimate pedagogy — that's not a calculator, that's a worked-example library. Animated tutorials in Plus add another layer that Wolfram Alpha never bothered building. At $69.99/year with a free tier that already delivers step-by-step explanations, the access economics are strong for individual learners globally.

The instructional architecture, though, stops at the student's phone. No LMS hooks, no API, no progress reporting, no instructor dashboard. If I'm a Head of Learning deploying this at scale, I have zero visibility into whether students are skipping steps or just screenshotting answers. That's not a minor gap — that's the difference between a learning tool and a homework shortcut.

If we adopt this for supplemental self-study support, it holds for 3 years. If we try to build it into a structured curriculum with assessments and learning analytics, we'll hit the wall fast. Socratic by Google had the same ceiling before Google folded it into Lens. Neither was designed for institutional learning architecture.

$9.99/month or $69.99/year, with a 6-month plan also offered in the app. No add-on modules except Plus unlocks word problem NLP and animated tutorials. Free plan includes step-by-step explanations — rare generosity at the base tier. Wolfram|Alpha also charges a subscription for step-by-step solutions; Photomath wins on camera-scan UX and support for 32 languages.

TCO math is simple: $69.99/year × 3 years = $209.97. No seats, no overages, no API costs — this is consumer software. A school district buying 500 licenses changes the math entirely, but no published volume pricing exists. That's the real gap: institutional buyers hit a wall.

Contract terms are published in the help center: Plus renews automatically at the end of each monthly, 6-month or yearly period, and you cancel in the App Store or Google Play at least a day before the period ends. No 30-day notice trap. Individual buyers face minimal risk at $70/year. Procurement teams buying at scale should get terms in writing.

The core workflow is tight. Point, scan, get steps. Handwriting OCR plus multi-method solving — elimination method, Cramer's rule, comparison method — means students aren't locked into one approach. That's the kind of scaffolding that maps to how I'd sequence a lesson. Solving runs in the cloud, though, so patchy classroom Wi-Fi stalls it. At $69.99/year, it's cheaper than one tutoring hour.

Day three is where trainer concerns surface. Students who reach for the camera before attempting a problem are practicing retrieval of Photomath's steps, not their own reasoning. Wolfram Alpha has the same problem, but at least it feels like a reference tool. Photomath's instant-solve UX is frictionless by design — which is exactly backwards for skill-building.

Word problem NLP is Plus-only, and documentation beyond the Google help center is thin — no changelog, and the articles blog is written for students and parents. Animated tutorials show promise for independent review between sessions, but without a teacher-facing dashboard, there's no way to track whether students are engaging with the 'why' explanations or just screenshot-collecting answers.

Camera recognition on handwritten math — that's the trick most apps fumble. Photomath's computer vision plus OCR combo reads messy student handwriting and spits back step-by-step solutions in real time. Multi-method solving is the standout: showing Cramer's rule AND elimination for the same system of equations is genuinely useful for someone trying to understand a concept, not just copy an answer. Compare that to free Wolfram Alpha, which gives you the answer and keeps step-by-step solutions for its paid Pro plan.

The free tier is honest. Step-by-step explanations, visual aids, 'how and why' tips — that's not a locked demo, that's a real product. Word problem support and animated tutorials sit behind Plus, which is the real tradeoff: NLP-powered word problems feel like a daily homework feature, not a premium luxury.

The big catch is that this is mobile-only by design. No web app, no desktop. That's fine for scanning homework on the kitchen table, but the moment a student is at a laptop doing practice problems, they're switching devices. Support for 32 languages is genuinely thoughtful, though solving runs in the cloud, so no connection means no answers. The camera dependency just makes the whole thing phone-shaped.

Three tells from the website. One: meta description calls it 'top-rated' — the kind of superlative that ages poorly. Two: a help center and an articles blog, but no changelog or API docs. Three: ownership sits in the fine print — the About page gives Photomath's 2023 move to Google one sentence, and the Terms name Photomath LLC at Google's Mountain View address.

What's actually working: the free tier delivers step-by-step explanations at $0, and the Plus annual plan lands at $5.83/month — honest value for a student budget. Multi-method solving (Cramer's rule vs. elimination, for example) is a real differentiator over tools that give you one path. Support for 32 languages is category-rare.

The exit story is bleak. No API, no web app, no data export story visible. If Photomath shifts direction — it's already been acquired once — you're copying screenshots. That's the real risk. Not the product quality. The dependency.