Basis Peak vs Fitbit Charge HR: the 2015 head-to-head, revisited

In late 2014 and early 2015, if you wanted a wrist tracker that read your heart rate all day, you had two real options. Intel’s Basis Peak shipped in November 2014 at $199. Fitbit’s Charge HR followed in January 2015 at $149. Almost every review of the period pitted them against each other, and almost every shopper choosing between them ended up with the Fitbit.

The Peak was the more interesting device on paper. The Charge HR was the one that sold.

This is the comparison that defined the category for a year, and it’s worth doing properly now that the dust has settled. One of these products is a museum piece. The other became a franchise that is still being updated, in modified form, in 2026.

The specs, side by side

The price gap is the first thing to notice. Fifty dollars, in the $150-$200 tracker market of 2015, was material. The Peak was being sold as the premium option against a band that did most of what shoppers cared about for two-thirds the money.

Form factor: a watch that wanted to be a band

The Peak was a watch. A small one, by 2014 standards, but a watch: square stainless case, sapphire-coated glass over a 1.25-inch monochrome OLED, swappable straps. You could read the time on it across a room. The touchscreen, the swipe gestures, the always-readable face: this was Basis trying to compete with the Apple Watch that everyone knew was coming in April 2015.

The Charge HR was a band. A textured plastic strap with a tiny OLED window that showed one line of text at a time: time, then steps, then heart rate, by tapping or pressing the side button. The display was the smallest reasonable thing you could put on a wrist and still call it a screen. Fitbit’s bet was that most people wanted the data on their phone and a minimal reminder on their wrist.

Both bets had merit. The Peak’s screen was more useful at a glance. The Charge HR’s was less obtrusive and looked more at home with office wear. In contemporary reviews, the Peak picked up “feels like a real watch” praise and “feels chunky” complaints in equal measure, while the Charge HR got “almost forget you’re wearing it” and “the screen is too small to read while running.”

In retrospect this was the wrong axis to fight on. The Apple Watch arrived five months after the Peak, with a screen the Peak couldn’t approach, and absorbed all the “I want a real watch with sensors” demand. The Charge HR sat below the Apple Watch on price and weight and ate the rest of the market.

Sensors: the Peak’s real advantage

This is where the Peak was honestly ahead.

Both devices used optical photoplethysmography (PPG) on the wrist for heart rate: green LEDs through the skin, photodiode catching the reflected pulse. Both were continuous, meaning the sensor sampled around the clock rather than only during a tagged workout. That part was a tie.

What the Peak added: a skin-temperature thermistor on the case back, and a pair of galvanic skin response (GSR) electrodes that measured perspiration as a change in skin conductivity. The Charge HR had a 3-axis accelerometer and nothing else.

The extra sensors mattered because Basis’s whole product theory was that sleep, recovery, and metabolic state are not legible from movement and pulse alone. Skin temperature drops at sleep onset; sweat picks up under cognitive or physical load; the combination of HR, motion, GSR, and temperature gave the Basis app something to model. We have a longer piece on the Peak’s sleep tracking that goes into the inference logic, because it was the part of the device that most clearly anticipated Oura.

The Charge HR, by contrast, was honest about what it had. Fitbit’s app showed steps, floors, distance, calories, resting heart rate, and a movement-based sleep estimate. That was it. No claim to read your stress, no skin-temperature trend, no perspiration data.

Battery: a near tie, with caveats

Intel claimed four days for the Peak. Fitbit claimed five for the Charge HR. In normal use we and most contemporary reviewers found both numbers slightly optimistic and the gap not as wide as the spec sheet suggested — call it three to four days for the Peak, four to five for the Charge HR.

The Peak’s screen was bigger and the optical sensor sampled more aggressively to feed the multi-signal sleep model. Both cost battery. The Charge HR’s tap-only display and accelerometer-led tracking sipped less.

Either was a long way ahead of an Apple Watch, which needed a charge every night. Neither was the reason anyone picked one over the other.

Accuracy during workouts

The accuracy story split cleanly by activity, and it was reported consistently in 2015 reviews from DC Rainmaker, The Verge, Wareable, and Engadget.

At rest and during sleep, the Peak’s optical heart rate was the steadier reading. A few users found it consistently within a couple of beats per minute of a chest strap during sedentary measurement, where the Charge HR drifted more. Basis had spent two product generations tuning that sensor for low-motion conditions.

During high-intensity intervals — HIIT, sprints, heavy lifting, anything with sharp wrist motion — the Peak fell apart faster than the Charge HR. The signal dropped, the watch displayed dashes, and the recorded workout was missing chunks of data. The Charge HR, with a smaller and more aggressively filtered optical stack, would more often produce a number, even when that number lagged the actual heart rate by ten to twenty seconds.

Neither device was a substitute for a chest strap, and neither was FDA-cleared for anything. We cover this in detail in the Peak’s heart-rate accuracy archive. The short version: if you cared about resting and sleeping heart rate, the Peak read better. If you cared about lifting and intervals, the Charge HR was less frustrating.

Apps and platforms: where Fitbit was already winning

The Basis app and the Fitbit app were aimed at different people, and Fitbit had picked the larger group.

Basis built Habits and Insights. Habits were behavioral goals: “wear it more than 12 hours a day,” “go to bed by midnight three nights this week,” “walk 30 minutes before noon.” You unlocked them gradually so the system didn’t drown you on day one. Insights were generated weekly: trends in sleep, in resting HR, in skin-temperature variability. The app rewarded quiet, repeated behavior. It was good. It also assumed you were already interested in your own physiology.

Fitbit built badges, leaderboards, and a friends feed. You got a badge for your first 10,000-step day, another at 25,000 lifetime miles, a “Sahara” badge at 2,983 miles. You could see your sister-in-law’s step count and pass her on the weekly leaderboard. The app rewarded loud, social behavior. It was less sophisticated about the body. It was wildly more effective at getting people to keep wearing the device.

Fitbit also had distribution that Basis did not. By the end of 2015, Fitbit was the number-one wearable brand worldwide by units, with retail presence in Best Buy, Target, Costco, and most US drugstore chains. Basis sold through its own website and a handful of partners. We have a longer piece on why the Basis Peak failed that breaks down the commercial side of this.

App cadence is the part that gets least credit. Fitbit shipped meaningful app updates roughly every six to eight weeks across 2015 and 2016. Basis shipped them rarely, and after Intel’s acquisition the team’s engineering bandwidth went into the firmware that ultimately caused the recall. By mid-2016 the Basis app looked like it had in late 2014. The Fitbit app had been redesigned twice.

Why Fitbit survived

The recall is the obvious answer and it isn’t the whole answer.

Intel bought Basis Science in March 2014 for a reported $100M–$150M and put it inside the New Devices Group, a division whose remit was unclear from the start and which Intel itself wound down in 2017. The Peak was Basis’s first and only post-acquisition product. Internal accounts from former employees, published in The Information and reported on at the time, describe a small team trying to ship hardware on Intel’s timelines while losing autonomy over the roadmap. When the November 2015 firmware update started causing the overheating reports, the team didn’t have the latitude or the spare capacity to pull the update and start over.

Fitbit had four things Basis didn’t: a public market (it IPO’d in June 2015), a deep retail footprint, a product line wide enough to absorb a single SKU failure, and a CEO (James Park) who was running his own company. None of that guaranteed survival; Fitbit’s own stock collapsed from 2016 onward and Google eventually acquired it in 2021. It was still enough to outlast Basis by more than a decade.

The Peak’s June 2016 recall and the December 2016 cloud shutdown closed the case. Roughly 200,000 units, refunded; the team folded; the brand retired. Fitbit, meanwhile, shipped the Charge 2 in September 2016, the Alta HR in March 2017, and a Charge 3 in 2018. The line never paused.

What carried forward: from Charge HR to Charge 6

The interesting thing about the modern Fitbit Charge 6 is how much of it is recognizable as the Charge HR’s grandchild.

The form factor is still a slim band with a small color screen, worn around the clock, charged once or twice a week. The headline feature is still continuous wrist HR with workout tagging. The app is still organized around steps, sleep, and a daily readiness-style score (Fitbit calls it the Daily Readiness Score and gates it behind Fitbit Premium). The badges are still there.

The Charge 6 also picked up most of what the Peak had been right about. It has skin-temperature variation overnight. It has electrodermal activity for stress check-ins. It has SpO2. It has sleep stages inferred from multiple signals rather than motion alone. The model the Basis app shipped in 2014 — a wrist device that watches the body across several signals and quietly tells you something useful at the end of the week — is now the default expectation for a $160 fitness band.

Two things changed underneath. The first is that Fitbit Premium is now the business model: the deepest insights (the Readiness Score, the daily stress score, advanced sleep analytics) sit behind a $9.99/month subscription. The Charge HR’s data was all free. The second is that the device is now an arm of Google, with the Google account migration that came with that, which some long-time Fitbit users still resent.

For most people choosing a slim wrist tracker in 2026, the Charge 6 is the device the Peak was trying to be. Our fitness tracker guide covers the alternatives, including Garmin’s Vivosmart line for anyone who’d rather not pay a subscription to see their own data.

What the comparison says about the category

Buying a Peak in 2015 meant betting on the more thoughtful product. Buying a Charge HR meant betting on the company that was going to be around. Both bets were defensible at the time, and one of them turned out to be wrong, but not because the Peak was a worse device. It was the better-engineered piece of hardware, with a more interesting sensor stack and a more honest theory of what a wearable was for. It lost anyway, for reasons that had very little to do with what it could measure off your wrist.

The wearable market did not reward the most accurate sensors or the most defensible inference model. It rewarded the company that shipped app updates every six weeks and got its bands into Target. We try to keep that in mind when we look at the current crop.