Google Health continues to evolve, demonstrating progress despite its inherent complexities.
Upon engaging with the Google Health Coach, the system immediately assesses the user's physiological state, suggesting a potential for physical strain. My sleep metrics were deemed suboptimal, resulting in a low readiness score. Heart rate variability, an indicator of recovery, registered below my baseline. The coach also noted my extended exposure to a hot, humid environment, acknowledging ambient temperatures exceeding 90 degrees Fahrenheit. Consequently, Google's AI coach advised against planned strength workouts, prioritizing hydration, heat avoidance, and light activity. It even inquired about calf discomfort and my overall perception of its assessment.
My reaction to this evaluation was, frankly, mixed. This sentiment accurately encapsulates my experience over the past month while testing the $99 Fitbit Air. If the discussion were limited solely to the hardware, my sole critique would be the mislabeling of my "lavender" device, which is distinctly periwinkle. However, this review primarily focuses on the software experience and the growing influence of AI in consumer healthcare.
The Fitbit Air boasts exceptional battery longevity, a remarkably lightweight and comfortable design, and, significantly, offers basic tracking data without a paywall for users who prefer to bypass AI coaches. When utilized effectively, the integrated AI coach can prove genuinely beneficial, all at an accessible price point.
Conversely, the Google Health app requires further refinement, and its AI health coaches demand substantial user input to yield optimal results. The aforementioned color discrepancy also stands out as a minor, yet persistent, aesthetic issue.
As a foundational tracker, the Air excels in areas where Fitbit has traditionally shone. It's so lightweight it's often imperceptible on the wrist, offers extended battery life, and recharges rapidly. During WWDC, I received a 20 percent battery notification; a mere 45-minute charge while preparing for the day boosted it to 85 percent, negating any further battery concerns since. In approximately a month of use, I've only needed to charge the Air three times. While the reliance on a proprietary charger is a minor inconvenience, it's a common characteristic among fitness bands.
While it may not offer the absolute broadest spectrum of metrics available, the Air provides a comprehensive dataset sufficient for all but the most dedicated health enthusiasts. Core functionalities include step count, resting heart rate, and sleep tracking. More advanced metrics encompass heart rate variability, blood oxygen levels, readiness scores, detailed sleep stages, and cardio load, which Fitbit uses to contextualize optimal weekly cardiovascular activity. The device omits push notifications but effectively supports silent alarms for morning wake-ups.
Compared to a Whoop band, the Air's default textile strap is noticeably thinner, sleeker, and simpler to secure. (The Whoop clasp has been a persistent annoyance for me.) The sensor module can be easily detached and reinserted into different straps, allowing for stylistic changes. While Google's official alternative straps aren't a favorite, the release of specs and guidelines suggests a promising future for third-party options. Aesthetically, the Air's most significant advantage is its resemblance to a stylish bracelet. For those with smaller wrists, the Air accommodates sizes from 130mm to 210mm, though wrists around 5.75 inches (146mm) feel like the lower limit before potential sizing awkwardness, such as gaps or excess strap material, becomes apparent.
As previously stated, my hardware-related complaints are virtually nonexistent. While a $99 annual Google Health Premium subscription exists, it's entirely optional. This subscription unlocks a video workout library, adaptive fitness plans, more in-depth metrics, and access to the AI health coach. Crucially, basic fitness tracking data is no longer paywalled. However, Google anticipates users will seek insights beyond raw numbers, which is precisely where Google Health and its AI coach aim to deliver.
My long-held view on consumer-focused AI health and fitness features, as I've frequently articulated, is that they often represent poorly integrated, opportunistic attempts that offer little beyond readily available information. The promise of personalized health insights driven by individual data often falls short of reality. Nevertheless, among all such systems I've evaluated, the Google Health Coach comes closest to being genuinely useful, provided one is willing to invest significant effort.
The Google Health Coach, powered by Gemini, doesn't revolutionize AI fitness but is now a central feature within the rebranded Google Health app. Each morning, it provides a summary of sleep and readiness metrics, followed by daily recommendations. It can address health-related queries, interpret data trends, and suggest adjustments to fitness plans. For instance, facing a series of business trips and medication side effects, the coach successfully generated a travel-friendly workout routine with a reduced step goal and bodyweight strength exercises. It consistently defers to medical professionals for diagnoses and advises consultation with healthcare providers.
The Health Coach is not exclusive to the Air; Pixel Watches also incorporate it, with future plans for expansion to third-party wearables. Since October 2025, nearly 500,000 individuals have beta-tested the feature, generating over a million feedback points that informed an improved version released last month. I can confirm noticeable enhancements, including a more customizable (though still somewhat cluttered) layout, the introduction of leaderboards for social competition, and a chatbot that is approximately 30 percent less verbose. It now provides sources for health facts—many of which, in my testing, were reputable clinical studies or sources—and allows for medical record uploads to provide additional context for analysis. Note that uploading records necessitates identity verification via CLEAR, the same service used for airport security, and requires periodic permission renewals.
Upon morning synchronization, the coach delivers a concise analysis of readiness and sleep scores, followed by actionable advice. For example, during recent dehydration and fatigue while covering the Enhanced Games in Las Vegas, the coach recommended hydrating with electrolytes, skipping strength workouts, reducing step count, and prioritizing rest. Similar advice was provided days earlier at Google I/O and again during WWDC. It also suggested bland foods like bananas, rice, and applesauce to prevent gastrointestinal upset and nausea.
While these recommendations were beneficial for my health, they were a direct result of an extensive initial setup process. I dedicated five to six hours to meticulously informing the coach of my specific needs, outlining my three-, six-, nine-, and twelve-month goals, and painstakingly uploading and explaining roughly ten years of medical history. I also enabled every conceivable data integration a wearable reviewer might utilize. I detailed all my medications, dosages, and diagnoses, and manually entered three sets of recent blood test results where direct medical record uploads were not feasible due to less tech-savvy healthcare providers. Regrettably, the Google Health Coach cannot interpret screenshots. Even after this exhaustive effort, the coach occasionally reverted to older data during subsequent check-ins, necessitating reminders of previous conversations.
This process proved exceedingly tedious, particularly when the coach appeared to lose context. Other inconsistencies also emerged. When I initially configured the coach in October, I had not yet started my current medication regimen, maintaining a 10,000-step daily goal, running three to four times weekly, and lifting two to three times weekly. However, since January, my physical capacity has been limited to walks and light strength training. The Google Health Coach commendably adapted by crafting fitness plans that prioritized muscle retention and set a more manageable 5,000-step daily goal, accounting for my fatigue. Yet, while some sections of the app reflected this change, others persistently displayed my original 10,000-step goal, an inconsistency the AI failed to permanently correct despite repeated attempts.
As a highly self-quantified individual adhering to a specific treatment plan, I undergo quarterly blood tests, biannual ultrasounds, and monthly doctor follow-ups. My meticulousness stems from significant physiological changes and side effects. However, what about users who are not as diligent? What if there's no incentive to dedicate several hours weekly to educating, updating, and verifying the health coach's information? Such users are likely to encounter a less intelligent AI coach and a significantly less valuable experience.
Out of curiosity, I surveyed several fellow tech journalists testing the Air and Google Health app, and their experiences varied considerably. My colleague David Pierce reported a useful interaction; when severely unwell, the AI coach correctly urged him to seek hospital care. Conversely, our senior news editor, Richard Lawler, expressed strong dislike for the feature, wishing it could be permanently disabled. Another reviewer described the coach as "Bad" with a capital B, noting they only interacted with it daily and provided no medical data. Yet another found its insights no better than common sense. However, some praised the ability to log nutrition by photographing meals and then refining entries through chatbot conversation. Overall, the quality of each individual's experience directly correlated with their patience for interacting with the AI.
For me, the AI coach's most effective current application is as a supportive tool between doctor visits. I input my doctor's recommendations and leverage the coach for accountability in adhering to that advice. This involves daily check-ins to ensure I'm prioritizing protein in meals and focusing on strength training and low-impact cardio. It aids in monitoring progress and conducting preliminary research into potential causes of side effect flare-ups. Before appointments, I can prompt it to review my recent history and formulate follow-up questions for my doctor based on the previous month's data.
The ingenuity of the Fitbit Air lies in its capacity to offer two distinct software experiences on the same hardware. For those averse to AI who simply desire a traditional fitness tracker, the $99 hardware purchase grants access to free, basic data tracking, with no obligation to interact with the AI coach. However, for users who wish to engage with the AI, that experience also begins with a $99 upfront cost, which includes a three-month premium subscription trial.
Google Health remains an ongoing project. Yet, the Air stands out as Google's most astute wearable release since the Fitbit acquisition. There's a clear demand for simpler trackers that still incorporate advancements in health tech software. By combining an affordable price with the option to bypass the AI chatbot, Google has created a device that appeals to both AI enthusiasts and skeptics alike. Achieving such a balance in the health tech sector is an increasingly challenging feat.
Every smart device now mandates agreement to numerous terms and conditions—contracts rarely read by users. While it's impractical for us to meticulously analyze each agreement, we now tally the number of "agree" prompts required for device usage in our reviews, recognizing that these are agreements most people don't read and cannot negotiate.
To operate the Google Fitbit Air, pairing with an iPhone or Android smartphone is essential. This entails agreeing to the phone's Terms of Service, privacy policy, and any granted permissions. The same applies to any third-party health integrations. For example, opting to upload medical records requires agreeing to use CLEAR for identity verification, thus also accepting CLEAR's terms of service and privacy policies. Additionally, there are several optional permissions for features such as location, Bluetooth, camera, background app refresh, notifications, and cellular data.
By configuring the Fitbit Air, you are consenting to Google's Terms of Service and Privacy Policy, Fitbit's Terms of Service, and Google Health's Terms of Service.
For users concerned about their health data being used for Google's AI training, this option is deactivated by default and requires explicit opt-in. Furthermore, the terms established during Google's acquisition of Fitbit stipulate that your health data must be stored separately and cannot be leveraged for Google's targeted advertising business.
Final Tally: Three mandatory agreements, several optional permissions.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.
