How to choose a CBT-I app for chronic insomnia

For adults with chronic insomnia, a CBT-I app is worth considering when the alternative is no CBT-I at all: a long therapist waitlist, no trained clinician nearby, or a schedule that makes weekly appointments impossible. That does not make every sleep app a treatment. The useful first question is blunt: does the program actually deliver cognitive behavioral therapy for insomnia, including the uncomfortable parts, or does it mostly track sleep and play calming audio?

If you are still confirming whether your sleep problem meets the threshold for chronic insomnia, start with chronic insomnia disorder. If you already know CBT-I is the recommended behavioral treatment but cannot get it in person, the access gap is the reason these platforms matter; it is also why weak apps should not get a free pass. The better comparison is not “Which app looks polished?” but “Which one gives me the full protocol, has published evidence, fits my risk profile, and offers enough support that I might finish it?”

Start With The Five Parts, Not The App Store Rating

A full CBT-I program includes sleep restriction, stimulus control, cognitive restructuring, sleep hygiene, and relapse prevention. Sleep hygiene alone is not CBT-I. A diary alone is not CBT-I. Relaxation recordings may be pleasant, but they do not replace the behavioral work that makes CBT-I different from general wellness advice.

The American Academy of Sleep Medicine’s digital CBT-I platform characteristics table is useful because it compares platforms on the things a patient cannot reliably infer from screenshots: published randomized controlled trials, program length, FDA status, coaching model, and cost or access route. In that table, Sleepio has 12 published RCTs, Somryst/SHUTi has 15, CBT-i Coach has 3, and several newer programs have 0 to 1 or no published RCTs listed; Somryst is identified as the only FDA-cleared option for chronic insomnia.[1]

That table should not be read as a beauty contest. It is a triage tool. Somryst/SHUTi and Sleepio deserve more attention because they have the strongest trial portfolios. CBT-i Coach deserves a different kind of respect: it is free and can be useful, especially for someone already working through CBT-I concepts, but it is not the same thing as enrolling in a full guided or automated treatment course. Newer commercial programs may still help some people, especially if coaching improves follow-through, but the evidence burden has not disappeared just because the interface feels friendlier.

What The Evidence Says Digital CBT-I Can Do

Digital CBT-I is not just placebo-like sleep content. A 2025 meta-analysis of 49 randomized controlled trials with 20,118 participants found that digital CBT-I reduced Insomnia Severity Index scores more than control conditions, with a weighted mean difference of −3.42 points; in studies using a stricter insomnia threshold of ISI ≥ 10, the reduction was larger at −4.25 points, and U.S.-based studies showed a weighted mean difference of −4.02 points.[2]

Those are average effects, not a guarantee that one exhausted person will sleep normally after a fixed number of weeks. An ISI drop can be clinically meaningful without making insomnia vanish. It also matters that the larger effects appeared in samples more clearly selected for insomnia. If a person has occasional stress-related poor sleep, a full CBT-I app may be more treatment than they need. If a person meets chronic insomnia criteria, the evidence is more relevant.

The boundary is just as important. A 2023 network meta-analysis found stronger effects for in-person or telehealth CBT-I than for digital delivery.[3] That does not make digital CBT-I useless; it puts it in its proper place. When trained CBT-I care is available and feasible, especially for complicated cases, digital treatment is not automatically the superior option. When trained care is unavailable, digital CBT-I can be a serious substitute for waiting indefinitely.

A 2026 review reported that FDA-cleared apps show sleep efficiency gains of 10% to 15% and ISI drops of 4 to 7 points, while consumer sleep apps are more inconsistent.[4] The same review flagged a familiar problem: automated programs can lose 30% to 40% of users after 4 weeks.[4] That is not a minor usability footnote. CBT-I asks people to change the timing of sleep, get out of bed when awake, and tolerate temporary restriction of time in bed. A program can be evidence-based and still fail a person who cannot stick with it.

Somryst/SHUTi And Sleepio Are The Evidence Anchors

Somryst/SHUTi stands out for two reasons that are easy to confuse but should be kept separate. First, the SHUTi program has a large published trial history, with 15 RCTs listed in the AASM platform table.[1] Second, Somryst is the FDA-cleared prescription digital therapeutic for chronic insomnia in that comparison.[1] FDA clearance is not the only standard worth caring about, but it is a meaningful regulatory signal when the choice is between a treatment platform and an app-store product borrowing clinical vocabulary.

The tradeoff is access. A prescription digital therapeutic can be harder to obtain than a direct-to-consumer app. Coverage and out-of-pocket cost vary, and a clinician may need to be involved. For some patients, that friction is acceptable because the platform’s evidence and regulatory status are exactly what they want. For others, especially those paying cash or trying to start quickly, the access route may be a real barrier.

Sleepio is the other major evidence anchor. The AASM table lists 12 published RCTs for Sleepio, placing it well ahead of most consumer-facing options.[1] It is not the same regulatory category as Somryst in the AASM comparison, but published RCT support matters. For an adult with uncomplicated chronic insomnia who wants a structured automated program and can access Sleepio through an employer, health plan, or available consumer pathway, it belongs on the serious shortlist.

The shared limitation of these stronger platforms is not that they are “too clinical.” It is that clinical programs are often built around eligibility assumptions. Many digital CBT-I trials and platforms exclude or caution against use in people with moderate-to-severe depression, bipolar disorder, schizophrenia, shift work sleep disorder, pregnancy, or complex medical situations.[1] Sleepio is described as a partial exception in the available comparison, but the larger point remains: a person with complicated insomnia should not be waved into an app as if exclusion criteria are paperwork trivia.[1]

Where CBT-i Coach Fits

CBT-i Coach is appealing for the obvious reason: it is free. It can help with sleep diaries, reminders, education, and practicing CBT-I skills. For someone already seeing a clinician, reading a self-directed CBT-I guide, or trying to organize the behavioral pieces, that is useful support. The AASM table lists 3 published RCTs for CBT-i Coach, which is more than many wellness apps can claim.[1]

The mistake is treating “free CBT-I app” as meaning “complete stand-alone CBT-I course.” CBT-i Coach was designed more as a companion tool than as a full commercial therapeutic pathway. If you need the app to calculate a sleep window, advance lessons, press you through stimulus control, and help you manage relapse, check whether the actual workflow does those things rather than assuming the label covers them.

For readers who want to try the behavioral method manually before paying, a self-directed starting point such as a CBT-I home approach for insomnia may pair more honestly with CBT-i Coach than pretending the app alone is the treatment.

Coaching May Matter More Than The Sales Page Admits

Human coaching is not automatically better science, but it can change completion. Guided programs show roughly 60% to 80% adherence, compared with 30% to 50% for fully automated programs. That gap matters because the hard parts of CBT-I arrive after the purchase, not before it.

Sleep restriction is the obvious stress test. A person may like the idea of evidence-based insomnia care until the program asks them to delay bedtime, get up at a fixed time, and stop extending the sleep window after a bad night. Stimulus control can be just as irritating: leaving the bed when awake sounds simple until it happens at 3 a.m. in a cold room. Coaching can help some users stay with these steps long enough to benefit.

That is where coached commercial programs such as Sleep Reset may be attractive, even if their published RCT portfolio is not as deep as Sleepio or Somryst/SHUTi in the AASM comparison.[1] A coached program is not automatically more effective than a well-tested automated one. The practical question is whether you are the kind of patient who completes structured self-guided work or the kind who needs a person checking in when the treatment becomes inconvenient.

Older Adults Should Not Be Dismissed, But Exclusions Still Matter

One lazy assumption deserves to be retired: older adults cannot use digital CBT-I. The SHUTi OASIS randomized trial enrolled 311 adults aged 55 to 95 and found that a fully automated digital CBT-I program produced large Insomnia Severity Index improvements, with effect sizes from d = −1.04 to −1.29 sustained at 12 months; 63% of participants completed the program.[5]

That trial is encouraging because it separates age from ability. A 70-year-old who is comfortable online and has uncomplicated chronic insomnia may be a better candidate for digital CBT-I than a 35-year-old with severe depression, rotating night shifts, and escalating sedative use. The match depends less on age than on safety, diagnosis, support needs, and whether the program’s assumptions fit the person using it.

The exclusion issue should stay near the front of the decision, not in fine print. If insomnia is occurring alongside moderate-to-severe depression, bipolar disorder, schizophrenia, pregnancy, shift work sleep disorder, untreated sleep apnea symptoms, unstable medical illness, or medication changes that require supervision, an app may still play a role, but it should not be the whole plan. This is where the broader 2026 insomnia treatment discussion, including CBT-I versus medication or combination treatment, becomes more relevant than a product comparison.

Cost Comes After Fit

Cost matters, especially when the alternative is paying out of pocket for specialty care. But the cheapest option is not automatically the best option if it lacks the components or support you need. A free tool that you do not know how to use is not a bargain. A polished subscription that avoids sleep restriction is not CBT-I. A prescription product that you cannot access or afford may be clinically strong but practically unavailable.

The available cost picture is uneven. CBT-i Coach is free. Sleep Reset has been described in the available materials as roughly $230–300. Somryst is insurance-dependent. Commercial comparison pages put in-person CBT-I courses around $900–2,400 or more, but those figures should be treated as directional rather than neutral pricing data because the cited source is commercial.[1]

A cleaner way to think about money is to ask what the purchase buys: a complete protocol, an evidence base, regulatory review if relevant, coaching or automation, and a realistic path through the hard weeks. If a lower-cost option gives you those things, good. If it gives you only tracking, graphs, and a soothing library, it belongs in a different category.

A Practical Matching Framework

There is no single winner for every adult with chronic insomnia. There are better and worse matches.

• If you want the strongest regulatory signal and can access prescription digital treatment, look first at Somryst/SHUTi, while checking eligibility and coverage.
• If you want a robust automated program with substantial published RCT support and have access through a plan, employer, or available consumer route, Sleepio is a serious candidate.
• If you need a free tool to support sleep diaries, reminders, and CBT-I skills, CBT-i Coach is useful, but do not confuse it with a complete stand-alone course.
• If you know you abandon self-guided programs, a coached option may be worth considering even when its RCT portfolio is thinner, provided it truly includes the full CBT-I protocol.
• If you have major psychiatric symptoms, shift work disorder, pregnancy, complex medical issues, or possibly another sleep disorder, use product comparisons only after clinical guidance.

The features to discount are the ones that dominate app stores: streaks, dashboards, relaxing sound libraries, generic sleep scores, and broad claims about improving wellness. They may make an app more pleasant. They do not answer whether the app delivers sleep restriction, stimulus control, cognitive work, sleep hygiene, relapse prevention, and enough structure to complete them.

If the hard part of your chosen program is sleep restriction, pause before quitting and troubleshoot the method itself; many people struggle there first. A guide to sleep restriction therapy not working is more useful at that point than downloading a second app that promises comfort without the behavioral work.

Choose in this order: full CBT-I protocol first, published evidence and FDA status second, support model third, cost and access fourth. If you are still deciding whether digital care is the right category at all, review why CBT-I is so hard to get, the 2026 guidance on nighttime sleeping problems, or a basic CBT-I FAQ before paying for a platform.

References

1. Digital CBT-I Platform Characteristics Table, American Academy of Sleep Medicine.
2. 2025 Meta-Analysis of 49 RCTs, PMC.
3. Network Meta-Analysis, Scientific Reports, 2023.
4. APA 2026 Review, Psychiatry Advisor, 2026.
5. SHUTi OASIS RCT, npj Digital Medicine, 2025.