Everything parents want to know before their child meets Terry.
Your child opens the app and meets Terry — a warm, funny dinosaur who loves a good chat. Terry starts with a casual conversation: asking about their day, what they're into, maybe a silly story about Stomper or Cleo (Terry's friends). Once your child is comfortable, Terry introduces a short speaking activity — something like a breathing game, a silly word challenge, or a role-play about ordering food at a cafe. The session usually runs 10–15 minutes. Afterwards, you get a short note from Terry about what happened and one question to ask your child tonight.
Yes — and this matters more than it might seem. Most apps that claim to help with speech are built for articulation delays or language development. They're not designed for children who stutter. BraveVoice is built around how stuttering therapists actually think: reducing avoidance, effort-over-fluency, never rewarding smoothness at the cost of bravery. Disfluencies are normal here. Terry never corrects a stutter — ever. The goal is a child who feels brave about speaking, not a child who hides.
What your child gets from session one: a coach who never comments on disfluency, never praises "smooth" talking, and treats every attempt to speak as the win — through exercises like Story Builder, Take a Stand, Brave Word, and Walk Through It, all built around effort and courage rather than fluency. That's the whole product working for a child who stutters, right now, not a placeholder version of it.
On top of that, we're finishing three additional stutter-specific clinical techniques — voluntary stuttering, easy onset, cancellation practice — built around established stuttering-therapy frameworks. They're held back from release until a speech-language pathologist has formally reviewed them, because these are techniques real SLPs teach in session, not general confidence-building, and we won't ship that without a clinician's sign-off. They'll unlock as an upgrade to an already-complete experience, not as the missing piece of one.
Absolutely. BraveVoice is built for any child who goes quiet when the world gets loud — shy kids, anxious speakers, children who freeze in class or won't raise their hand, kids who can talk at home but shut down with strangers or in groups, children with selective mutism, and children who lost confidence after being laughed at or talked over. Terry is built to meet all of them where they are.
If you're not sure which describes your child, the parent dashboard has a simple setting that helps Terry adapt his approach — and "I'm not sure yet" is a perfectly valid starting point.
BraveVoice works well for children and teens from age 3 through the teen years. Terry has four different voices depending on age — a gentle Fred Rogers-style approach for the little ones (3–5), a calm older-sibling voice for middle childhood (5–9), a consulting, peer-adjacent tone for tweens (10–13), and an equal-partner tone for teens 14 and up. The app adapts automatically based on your child's age, so it doesn't age out of usefulness as they get older.
It isn't speech therapy and it doesn't replace it. A qualified SLP knows your child, can assess them in person, and can do things an AI can't. If your child sees an SLP, BraveVoice is designed to complement that work — consistent low-pressure speaking practice between sessions, in a space where there's no social stakes. If you're on a waiting list or can't yet access an SLP, BraveVoice gives your child something to do in the meantime rather than nothing.
Terry's approach is built around well-established stuttering-specific guidance to avoid the things that make stuttering harder for children: fluency-first praise, avoidance coaching, treating disfluency as failure. Terry never says "slow down," never praises a stutter-free sentence over a stuttered one, and never coaches a child to hide what's happening. Three advanced stuttering techniques — easy onset, voluntary stuttering practice, and cancellation — are locked in the current version. They address direct stuttering modification that works best with individualised clinical guidance; they're built into the platform and will be unlocked through your child's own SLP rather than offered universally. If Terry ever says anything that feels off, the parent dashboard has a feedback field that feeds directly into how Terry works with your child.
After every session you get: a short summary of what your child talked about, the exercise Terry introduced, a confidence score, any brave moments your child had, and one specific question to ask your child tonight. The question is the part parents say matters most — it gives you something to start a real conversation with rather than "how was it?"
The confidence score starts at 50 for every child — that's the neutral baseline, not a judgment. It moves up or down session to session based on engagement and effort, not fluency or how "good" your child sounded. The dashboard shows it as a plain-language band ("Just warming up," "Finding their voice," "On a roll") right alongside the number, so you're never left wondering whether 47 is good or bad — the label tells you.
Yes. We don't sell data. We don't share session content with third parties. Audio from your child's microphone is processed in real time for speech recognition and then discarded — we don't store audio recordings. The parent dashboard is PIN-protected so your child can't access it mid-session. We comply with COPPA (US) in full. Full details at bravevoice.me/privacy.html.
BraveVoice is $10/month or $100/year. There's a 5-session free trial — no card required to start. Founding family pricing is locked in for life as long as your subscription is active. We'll never raise the price on families who joined early.
Sessions run 10–15 minutes on average, though Terry follows the child's lead — a quiet day might be shorter, a chatty one might run longer. There's no timer and no pressure to reach a fixed length. Most families find 3–4 sessions per week is the sweet spot: frequent enough to build momentum, spacious enough that it never feels like homework. Daily is fine too. Terry adjusts to where your child is that day, so consistency matters more than frequency.
Terry is a large language model — the same kind of AI that powers tools like ChatGPT, but running a carefully designed character with strict guardrails, powered primarily by Groq. The conversation is generated in real time based on what your child says, the profile you set up, and a detailed set of instructions about how Terry should respond to children with different speaking challenges. On iOS devices, your child's voice is transcribed using Groq's Whisper service; on desktop browsers like Chrome, it's transcribed using the browser's own built-in speech recognition. Audio is not stored after transcription either way. Terry's voice is generated using OpenAI's text-to-speech. The session summary and tonight's question are generated separately after the session ends. Your child's data never trains any AI model.
That's useful information, honestly. Some children resist anything that feels like practice. Terry is built to be engaging, not instructional — but no app works for every child. If your child genuinely doesn't take to it after a few sessions, cancel anytime. No lock-in, no awkward calls to customer support. And if you want to tell us why it didn't work, that feedback makes it better for the next family.
Still have a question? Email us at hello@bravevoice.me — we read everything.