How the AEB Score works

The AEB Score (ALL Electric Bikes Score) is one number from 0 to 10 that every e-bike on this site gets. It's built from six things you can check on each bike's page: range, motor, brakes and build, how practical the bike is to live with, its current price, and the quality of the parts the maker lists. Same rules for every bike, so you can compare them at a glance. It isn't my opinion of the bike. That's in each review.

The six categories

Each category is scored 3 to 10, so the most basic spec still earns 3. The four spec categories use a straight-line scale, Price uses a smooth curve and Parts adds up points for named parts. The AEB Score is the weighted total.

Range uses my own tested miles on one charge when I measured them, because that beats any claim. Otherwise it uses battery size in watt-hours, which is measured the same way by every brand, unlike range claims. Motor and hills uses torque first, because torque is what you feel starting from a stop and climbing. Brakes and build rewards hydraulic brakes, suspension and a higher max load. Practicality is weight, compared within the bike's type, so a cargo or fat tire bike is judged against other heavy bikes, not against a 40 lb commuter. Price is scored on its own, so it doesn't count any spec twice. It uses the current price on the maker's own site, never Amazon, for the cheapest version of the model I reviewed. A sale price counts, and a sold-out bike keeps its listed price. A bike that's no longer sold uses its last known price (see below). Parts rewards name-brand parts the maker lists, explained next.

How Parts is scored

Two bikes can have the same battery, motor power and brakes on paper but very different parts. Parts gives credit for the named, better-quality parts a maker lists on its own site, its manual or its FAQ. Each bike can earn up to 10 points:

• Brakes, up to 2.5: name-brand hydraulic brakes (Shimano, SRAM, Tektro, Magura) 2.5, other named hydraulic brakes 2, hydraulic with no brand listed 1.75, named mechanical 1, unnamed mechanical 0.75, rim brakes 0.25.
• Gears, up to 2.5: higher-tier Shimano or SRAM groups, a Gates belt or an Enviolo hub 2.5, mid-tier groups like Shimano Altus or Acera 1.75, entry-level groups or a brand with no tier listed 1.25, unbranded or single speed 0.5.
• Motor maker, up to 1.5: Bosch, Shimano, Brose, TQ, Specialized or Yamaha 1.5, another named motor 1, no maker listed 0.25.
• Battery cells, up to 1.5: named cells (Samsung, LG, Panasonic, Molicel) or a named system battery 1.5, otherwise 0.25.
• Torque sensor, 1: a torque sensor 1, cadence or not listed 0.25.
• Tires, 0.5 for a named tire, and extras, 0.5 for a named fork, shock or display.

Parts = 3 + 0.7 × points, so a bike with all generic parts scores about 4.4 and a bike with every top part scores 10. Parts counts for 15% of the AEB Score, and the other five categories keep their same proportions in the other 85%.

Only listed parts count. If the maker doesn't name a part, it scores as generic, even if I mention a better part in a video or another site lists one, because you can't check it. No brand gets a bonus for its name, and my own opinion of the parts stays in the review. Each bike's score card lists the parts that were counted.

A worked example: Mokwheel Obsidian

The Mokwheel Obsidian, from the specs on its review page and its price on the maker's site:

Price: 3 + 7 ÷ (1 + ($1,799.99 ÷ $1,500)²) = 5.9. Parts: 3 + 0.7 × 5.25 = 6.675. Points are rounded here. The score adds the category scores as shown (one decimal), plus the exact Parts score, then rounds to one decimal.

Missing specs and price

Missing specs score 4.0. If a maker doesn't publish a number I need, that category gets 4.0. In Brakes & build, which averages brakes, suspension and max load, each part the maker doesn't list scores 4.0 inside the average. It isn't dropped and the other categories aren't re-weighted, so leaving a spec off never helps a bike. I fill gaps only from the maker's own pages and note the source and date on the bike's page; I never estimate a spec. Right now 39 of 127 bikes have at least one category at 4.0, counting Price.

How price is scored. Price = 3 + 7 ÷ (1 + (price ÷ $1,500)²). A $1,500 bike scores 6.5. Cheaper bikes score higher on a smooth curve that gets close to 10 but never flattens out, so a $500 bike (9.3) still beats a $750 one (8.6). Pricier bikes fall toward 3: $3,000 scores 4.4 and $4,500 scores 3.7. I checked every current price on the maker's own site on Oct 28, 2026, never on Amazon. The price, the site and the date are on each bike's score card.

Bikes no longer sold use their last known price. When the maker no longer sells the model I reviewed, or sells only a newer version, I use the last price I can find for that exact model, never the newer one's. I look on the maker's own pages first, then the maker's page in the Wayback Machine, then the price I gave in my review or video description, and never Amazon. The score card on each of these bikes says "last known price" with the date and links the source. That's 38 of 127 bikes right now; they're marked † in the table.

No price at all scores 4.0. If I can't find any price for the model I reviewed, Price scores 4.0, the same as a missing spec. Right now every bike on the site has a current or last known price, so none score 4.0 on Price.

Only real range tests count. My tested miles replace battery size only when I rode one charge and measured it. That's true for 7 bikes so far. Miles spread over weeks of riding, or mostly ridden without assist, aren't a range test, so those bikes use battery size.

What the score does not measure

• How a bike rides: handling, how natural the assist feels, motor noise.
• Fit: whether it suits your height.
• Build quality over time, support and warranty.

Those are in my reviews and videos. Use the score to narrow a list, then read the reviews to choose.

Every bike's score

All 127 bikes on the site are scored. Bikes tagged Spec profile are ones I haven't ridden; they get the same formula. Tap a column heading to sort, or sort all reviews by score.

* Not listed by the maker (or, under Price, no price found), so scored 4.0. ‡ Part of Brakes & build (brakes, suspension or max load) isn't listed by the maker, so that part scored 4.0 inside the average. † Price is the bike's last known price because it's no longer sold.

What changed

v3 (Oct 29, 2026)

• New Parts category (15%): points for the named brakes, gears, motor, battery cells, torque sensor, tires and fork or shock the maker lists. Parts the maker doesn't name score as generic.
• The five v2 categories keep their proportions in the other 85%: range 21.25%, motor and hills 12.75%, brakes and build 17%, practicality 17%, price 17%.
• Of the 125 reviewed bikes, 56 scores went up (by up to 0.8), 51 went down (by up to 0.4) and 18 didn't change. Bikes whose makers list name-brand parts went up; bikes whose makers list few parts went down.

v2 (Oct 28, 2026)

• Price is now a scored category (20%), using the current price on the maker's site, checked Oct 28, 2026. The separate Value chip is gone.
• Five categories: range (25%), motor and hills (15%), brakes and build (20%), practicality (20%), price (20%).
• Range uses my tested miles when I measured a full charge.
• Motor uses torque first, on a 40 to 120 Nm scale.
• Weight is judged within the bike's type.
• Bikes that are no longer sold are scored on their last known price (Oct 28, 2026 update).
• A missing spec or price scores 4.0 instead of being re-weighted away, so every bike gets a score.
• Fix, Oct 28, 2026: inside Brakes & build, a missing brakes, suspension or max load figure was being left out of the average, which could lift a bike's score. It now scores 4.0 like any other missing spec. 44 scores went down, by 0.1 to 0.8, and none went up.

FAQ

Is the AEB Score a review rating?

No. It's one number from 0 to 10 calculated the same way for every bike from its published specs, plus my own range test when I measured one. What I think of a bike is in the written review and the video.

Is price part of the score?

Yes. Price is one of six categories and counts for 17%. It's the current price on the maker's own site (never Amazon), or for a bike that's no longer sold, its last known price, with the source and date on each bike's page. Cheaper scores higher on a smooth curve: $750 scores 8.6, $1,500 scores 6.5 and $3,000 scores 4.4. If I can't find any price for the model I reviewed, Price scores 4.0.

What if the maker doesn't list a spec?

That category scores 4.0, a below-average mark. Brakes & build averages three parts (brakes, suspension and max load), and any part the maker doesn't list scores 4.0 inside that average. Nothing is dropped or re-weighted.

Can a brand pay for a better score?

No. The formula is the same for every bike, and the only inputs are published specs, the parts the maker lists, the maker's listed price and my own tests. No brand gets a bonus for its name.

Why does Parts only count parts the maker lists?

Because that's what you can check. If a maker names its brakes, gears, motor or battery cells on its own site, the bike gets credit for them. If it doesn't, that part scores as generic, even if I mention a better part in a video, because I can't show you where it's written down. My opinions of the parts are in the review, not the score.

How often do scores change?

Every time the site is built. If a spec or price is corrected or I measure a new range, the score updates in that build. If the formula changes, this page changes with it and gets a new version number.

Example badge: 7.8 AEB Scorehow it works