Apps for PPA, compared honestly
This page is for someone choosing a home-practice tool for primary progressive aphasia, probably in a hurry, probably tired. We make one of the tools described here. We have tried to write it so that you could use it to decide against us, because on several of these comparisons that is the right answer.
Why PPA changes the choice
Nearly every aphasia app on the market was built for stroke recovery. That matters in four ways.
- Gains are word-specific. Across the PPA studies, practice helps the words practised and spreads very little to others. So the word list is the product. An app that will not take your own words is an app that decides the whole thing for you.
- The direction of travel is different. Stroke apps assume improvement and unlock harder levels. In a progressive condition the useful design lets a list shrink and the cues get heavier without anybody being told they failed.
- The partner is part of the work. The strongest PPA trial in this field trained the person and their communication partner together.
- Pressure is the enemy. Word retrieval in aphasia measurably raises stress, and anxiety is common in PPA. Timers, streaks, scores and red crosses are the wrong stimulus, however motivating they look in a product demo.
Tactus Therapy: start here
Made by a speech-language pathologist in Vancouver, Tactus sells one-time purchases on iOS and Android, with a free Lite version of every app. Naming Therapy was listed at $29.99 and the Language Therapy 4-in-1 bundle at $89.99 in September 2026; a third-party review notes prices fluctuate and the suites go on sale.
What makes it the default recommendation for PPA is customisation. You can add your own photos from the camera or library, with your own text and your own recorded audio cues, and in the 4-in-1 bundle those custom photos carry across the naming, comprehension, reading and writing modules. You can switch individual words off and restrict practice to one- and two-syllable words. The cue hierarchy is complete — semantic, phonemic, sentence completion, written, spoken model — and the user chooses when and which cue to take, which is as close to errorless practice as anything on the market. No timers, adult imagery, works offline, and it emails a score report showing which cue unlocked each word, which gives the therapist something concrete.
The vendor's own article for clinicians on PPA tells them to practise priority vocabulary, teach scripts, and begin home practice from the first day, and names Language Therapy's custom word lists as the tool for it. That is a vendor saying the right thing about a hard condition.
On evidence: Tactus cites a peer-reviewed study of the Language Therapy 4-in-1 suite used 20 minutes a day for four weeks in chronic aphasia. That study is post-stroke, not PPA, and independent reviewers note the research backing applies to that suite rather than to the Advanced apps. Being clear about it: there is published research behind this product, and it is not PPA research.
Gaps worth knowing before you buy. There is no partner or coaching mode, no live clinician portal, and no spaced review — it presents lists rather than scheduling them. Reviewers' complaints are the upfront price, a customisation bug that has since been fixed, and an interface some find dated.
Constant Therapy: better if the therapist already uses it
A Boston University spinout, on iOS, Android and the web, sold by subscription. Listed at $29.99 a month or $299.99 a year on their pricing page in September 2026, with a 14-day trial; store listings show a spread of older and regional tiers, and it is HSA/FSA eligible.
This is the most-studied product in the field by a distance: Des Roches 2015, Godlove 2019, a JMIR dosage study in 2022, a 2023 long-term practice paper. In one of those, home users mastered tasks in fewer days than in-clinic users. It has 90-plus task types across more than a million exercises with adaptive difficulty, it is actively maintained, and it holds a 4.5 rating from over 4,000 ratings.
The clinician loop is the best in this comparison. The free Clinician app lets a speech therapist assign homework and see results remotely; caregivers get a web dashboard. Clinicians report it saves them substantial paperwork per patient.
Say it plainly: if the therapist is already on the Clinician app, this is probably the better choice for your household, whatever else is on this page. A therapist who can set the homework and read the results every week is worth more than a better-fitting feature set nobody is steering.
The honest reservations. It cannot take your own words or photos at all — the independent Aphasia Software Finder directory records that it cannot be author-customised by clinicians or users, and store reviewers say the same. All of its published research is post-stroke and TBI; no PPA trial was found. Its content is stroke-generic, which sits awkwardly against the PPA guidance that generalisation from generic drills is weak. Recurring complaints are the cost on a disability income, being "advertised as free" and then paywalled, the US accent, and a UK aphasia test group finding the illustrated guide childish.
Cuespeak: for pronunciation and scripts
Built by a UK speech and language therapist, Apple only — iPhone, iPad and Apple-silicon Macs. Subscription, listed in September 2026 in the range of $7.49 to $12.99 a month, $129.99 a year, or $399 for three years, with a 14-day full trial and a separate one-off clinician version.
Strong where Tactus is thin. Scripts let you enter a personal story and generate exercises from it, which matters because script training is one of the better-evidenced PPA techniques. Articulation videos break words down by syllable and phoneme. Wordfinder handles family names and personal vocabulary. Results show accuracy, speed and how much help was requested, and can be emailed to the clinician; there is a clinician portal. It rates 4.7 from 86 ratings, it is listed by the UK Stroke Association among its aphasia therapy apps, and it is described as NHS-recommended.
What it does not have: your own photographs. And it is British-built, so a US clinician's specific complaint is that the vocabulary and imagery do not always land for American users, though settings help. No published trial of the app itself exists. Take the free trial before you subscribe.
Lingraphica: the route to a funded device
Two different things share this name. TalkPath Therapy is free, on the web and iPad, with around 13,500 tasks and clinician-assigned plans and reports. Be warned that the iPad app was last updated in October 2020 and sits at 3.4 stars with reports of white-screen crashes and login loops; the web version is the live product. It has no custom content. SmallTalk is a set of 14 free iOS apps — video exercises and picture-phrase communication boards — which are communication aids rather than naming practice, and are not customisable.
The part that matters for PPA is the devices. Lingraphica's speech-generating devices are covered by Medicare as durable medical equipment after a speech therapy evaluation and a physician's order, with the patient paying the standard 20% after deductible; the company says many users have little to no cost. There is no public retail price. Their PPA page describes adding your own icons, words and phrases, a digital memory book, voice preservation, and care-partner support included with every trial.
If what your family actually needs is a communication device rather than practice, this is the funded path and no practice app substitutes for it. Ask the therapist about an evaluation. The clinical guidance for PPA is to introduce communication aids early, while the person can still learn to use them comfortably, rather than waiting until speech has gone.
The small one-time apps
Talk Around It, from Neuro Hero in Ireland, sells per themed edition — Home, Nature, Personal, USA voice variants — listed at $17.99 to $21.99 each in September 2026, one-time, with free trial versions and real voiceover. It offers semantic, phonemic, written and first-sound cues and progress graphs. The catch is size: roughly 100 to 115 pictures per edition. One reviewing therapist reported exhausting a pack's content in a single hour-long session. The store rating is low on a very small number of ratings, and it was last updated in August 2024.
iName It, from Smarty Ears, iPad only, listed at $14.99. Fifty household nouns photographed in five room scenes, with a sensible four-level cue hierarchy and reports broken down by cue type that export to the clinician. Fifty words is fifty words. No custom content, and no significant update since around 2022.
Both are cheap and neither is a mistake as a supplement. Neither is enough on its own for ongoing PPA practice, and neither takes your own material.
My Aphasia Coach
Subscription, iOS, Android and web, listed in September 2026 at $9.99 a month for the video library, $19.99 a month for the full set of around 50,000 exercises, or $199.99 a year, with a 14-day trial. Multimodal exercises combining reading, writing, listening and speaking, automatic difficulty, and a free portal for speech therapists to assign work and monitor it.
Two honest cautions. There is no peer-reviewed evidence for it, and the content is stroke-focused with minimal personalisation and no personal photos. The other is the reviews: recurring complaints about billing errors, difficulty cancelling, aggressive sales tactics, and login problems. Judge that for yourself, but go in with your eyes open and use the trial.
The free and the new
AphasiaGPT is free, built by a stroke survivor, launched in November 2025 and actively updated. It offers LLM conversation practice, scenario roleplay, a vocabulary library and a separate clinician portal. It has no published evaluation, no offline mode, few ratings so far, and it collects audio. For someone with PPA there is a specific concern worth naming: an open-ended chatbot that expects you to keep up is a source of pressure unless it is heavily constrained, and pressure is what makes word retrieval collapse.
Alfee is free with no account, built by a developer for his mother. Read, Listen and Match tasks in English, Spanish and Portuguese. Receptive only — no naming, no speech, no tracking, and a small amount of content. Worth mentioning mostly as a precedent: another family built their own.
Aphasia Therapy Online is a free website with no signup, built by a speech pathologist, with browser-based drills in a dozen languages and no personalisation or tracking. A reasonable zero-cost supplement.
Two more for the toolkit rather than for practice. Speech Assistant AAC, listed at $23.99 one-time on iOS and Android, is text-to-speech with your own categories and phrase buttons, useful for hard days if the person still reads well. Proloquo2Go is listed at $249.99 and is heavily customisable with personal photos, but it was designed for developmental AAC users and its symbol grid is a lot of cognitive load for someone with PPA. Treat it as a later-stage option rather than a first purchase.
Bungalow Software has been selling home aphasia practice for Windows since 1995, designed from the start to be used with a helper, with subscription pricing from $39 a month for its eighteen programs. The interface is of its era. Not a fit unless your household already lives on Windows desktops.
Anki: free, fully personal, and it needs you
Anki is open-source flashcard software with a spaced-repetition scheduler. It is not an aphasia app and makes no claim to be. It is on this page because it is the only free, open tool in this survey with a published PPA outcome attached to it: a case study in which a 72-year-old with the semantic variant built 591 personal flashcards — picture, description prompt, written word as the answer, with a show-the-answer-then-read-aloud rule — and practised around 30 minutes three or four times a week for 20 months, with a spouse monitoring. The report describes 139 words retained at the end.
That is one person, not a trial. But the ingredients are exactly what the PPA literature keeps pointing at: personal words, personal pictures, an errorless rule, a spouse involved, and spaced review over a long period. If someone in your family is comfortable with computers and willing to build the deck, Anki costs nothing and personalises completely.
What you give up: no cue hierarchy, no partner-coaching guidance, no aphasia-friendly interface, no recorded family voices unless you add them yourself, and a setup burden on a family that has plenty to do already.
Communication Bridge: the best PPA evidence here, and you cannot buy it
Communication Bridge-2, from Northwestern's Rogalski lab, is the strongest PPA evidence in this entire survey. A randomised trial of 95 PPA dyads — the person and their communication partner — over 12 months, combining video-chat sessions with a speech therapist and a custom web app for home practice. Participants averaged about 13.7 exercises across roughly four days a week. Dropout was under 10%. The arm that trained the partner, used personalised photos and set personal goals did better on goal attainment than the impairment-only arm, and did not show the decline in communication participation over the year that the control arm showed.
You cannot download it. The web app is a research tool, and there is no public release. A follow-on pilot (CB3) is testing an internet-based version aimed at wider implementation. The realistic route is to ask your speech therapist about enrolling in the trial or about a referral to Northwestern's Mesulam Center.
It is worth naming what that trial implies for everyone else: personalised word lists, scripts, partner training, a simple home-practice app, and weekly clinician contact. That combination is the one with published PPA outcomes, and no product you can buy today assembles all of it.
What none of them do well
Reading across the whole landscape, the same gaps appear repeatedly.
- No partner or coach mode anywhere. Not one of these apps gives the person sitting beside the learner any guidance about which cue to offer next or when to stop. Tactus Conversation Therapy supports up to four participants, which is the closest thing on the market, and it still gives the partner no coaching.
- Personal photos are close to absent. Only Tactus lets you add your own photos with your own cues, and there they sit alongside a generic set rather than being the whole content.
- No spaced review. These apps present lists. None of them schedules a word's return based on how much help it needed.
- Scores everywhere. Most commercial naming apps mark answers right or wrong and show the result. Nothing in this market defaults to a no-score, no-timer, errorless design.
- Nothing is built for decline. Levels go up. Lists grow. That is the wrong shape for a progressive condition.
- Longevity is a real risk. One major free product's iPad app went more than five years without an update. Subscription products can be discontinued. Ask yourself where your photos and recordings live if the company stops.
A word about speech recognition
If an app promises to listen and tell you whether the word was said correctly, be sceptical. Published work on automatic speech recognition with aphasic speech is consistently poor: Whisper-large-v3 evaluated on AphasiaBank produced a 38.4% word error rate for people with aphasia, error rates rise with severity, single-word naming is the hardest case of all because there is no surrounding sentence for the model to lean on, and apraxic speech is worse again. Whisper also hallucinates text, most often around pauses and disfluent audio — which describes exactly the speech it would be listening to here.
This is why most established apps use self-scoring or the partner's judgement instead: Tactus, Talk Around It and iName It all do. Cuespeak's recognition is opt-in. Constant Therapy records for clinician review rather than auto-scoring speech. Where recognition is used well, it answers the narrow question "was that the target word, or close enough?" rather than trying to transcribe.
Where Words at Home fits
We built this after the research above, for a household where one person has PPA and a spouse runs the practice. It starts empty: your photographs, your recorded voices, your word list. The cue cascade runs in a fixed order you can change, always ending with the person saying the word. The learner's screen has one photo, one word, and three big buttons — Help, Stop, Got it — with no timers, scores, percentages or red anywhere. A scheduler decides when each word returns based on how much help it needed. Everything is logged quietly for the adults and exports as CSV for the therapist. There is an optional speech check that answers credit, close, or not sure, and never says wrong; it can be left off.
The honest comparison. Constant Therapy has a substantial published evidence base and we have none. Tactus is a mature, well-reviewed product from a speech-language pathologist with years of customer support behind it, and ours is not. Cuespeak has articulation videos and script tooling we do not have. Lingraphica has devices your insurance may pay for. If any of those descriptions is what your family needs, buy that one.
What we have that they do not is the combination the PPA research keeps pointing to and the market has not assembled: personal photos and family voices as the entire content, a coach mode for the person sitting alongside, errorless practice as the rule rather than a setting, and spaced review built for holding on to words rather than acquiring new ones. Whether that combination helps anyone is a question a trial would have to answer, and no trial exists.
One more resource worth knowing, independent of all of us: the Aphasia Software Finder, run by the Tavistock Trust in the UK, is a directory with an aphasia-friendly page and a professional page for each app, listing cueing, customisation, feedback and price band. It is the neutral place to check anything on this page.