What it is in one line
An app store for voice-native apps that lives in your Mac's notch: describe what you want,
it generates the app on the spot, and you share a link that anyone can install with one click.
Who built it
VoiceOS is a product of WakoAI, Inc. (San Francisco), founded in 2023 by Jonah Daian (CEO)
and Kai Brokering (CTO), from Y Combinator's X25 (Spring 2025) batch. Brokering previously
built Wako.ai and NuPhone (an AI voicemail assistant), making VoiceOS his third bet on
voice.
A correction to the source data: the pool file lists Garry Tan as the builder. Public
records show Garry Tan is YC's CEO, not VoiceOS's founder or hunter — the the launch platform
hunter credit belongs to Jonah Daian and Gabe Perez, and the maker is WakoAI. The pool
entry likely mistook "YC-batch company" for "YC's leader."
Read: three consecutive bets on voice, with YC backing each time — this is not a hobby.
It is real money on the judgment that voice interaction becomes mainstream, and the App
Store step is a land grab for the future distribution channel.
What it actually does
- A store for voice apps → a marketplace living in the notch, stocked with
voice-native apps rather than traditional ones
- One-sentence app generation → type or say "build an app to help me journal" and the
app is generated on the spot
- Link-based distribution → every generated app becomes a shareable link; anyone can
install it in one click, no account gate
- Runs inside the VoiceOS runtime → apps sit on top of VoiceOS's cross-app voice
capabilities (dictation and Agent Mode)
VoiceOS itself has two entry capabilities: Dictation Mode (context-aware dictation
across apps) and Agent Mode (voice instructions that trigger multi-step workflows,
unveiled at the "VoiceOS Hackathon" in Tokyo on 2026-04-03).
What old behavior it replaces
Two old behaviors:
- The old install flow — downloading from an app store, registering an account,
granting permissions. This store collapses that to clicking a link.
- The old app-building flow — writing code, debugging, shipping. Saying one sentence
and getting a working app replaces the time-to-runnable from days to seconds.
Its more important role is identity: claiming the distribution slot before voice apps
themselves exist. The historical pattern is that when a new interaction medium appears,
the channel makes money before the apps do (the App Store for iPhone, Alexa Skills for
smart speakers). That is the position VoiceOS App Store is staking out.
Business model
VoiceOS itself is Freemium: a free tier with no credit card, and a Pro tier starting at
about $15/month billed annually (per the voiceos.com blog). The App Store's take rate is
not disclosed.
Read: today it is not making money from the store; the store's job is to give the
subscription a reason to exist via an ecosystem story. The real bet is that if voice apps
take off, store commissions and in-app transactions become a second growth curve. Classic
channel-first, rent-later play.
Hard numbers
- VoiceOS App Store launched on PH 2026-08-09: 326 upvotes, 19 comments, #2 Product of the Day
- VoiceOS's own monthly visits were roughly 21K (toolify, as of April 2026) — early stage
- YC X25 batch member; funding amount not disclosed
- Team size not disclosed (two named founders; headcount unknown)
Four-way read
| Dimension |
Call |
| Founder-product fit |
Three consecutive voice products from the same team; the conviction is consistent — among the best fits in this batch |
| Product insight |
Nailed the "the channel wins before the apps" pattern and is claiming the plot before the ecosystem exists |
| Execution quality |
Generating apps on the spot and shipping them as one-click links implies a real app template/runtime, not vaporware |
| Timing |
Half a step early. The store may spin empty for a while, but owning the slot early is the win |
The call
A classic channel land-grab, betting that the voice-app ecosystem will arrive.
To judge this, separate two things: VoiceOS itself (dictation + agent, genuinely useful,
already working) and the App Store (an ecosystem bet on a market that barely exists yet).
The first has proven out — it has users, a published price, and third-party praise. The
second is buying position for later.
The risk is legible: the store's fate does not depend on VoiceOS, it depends on
whether anyone builds voice apps worth installing. If in twelve months the store is still
indie demos, the slot spins empty. On the other hand, the fact that a YC company is
betting on voice-native apps suggests the category is probably not a phantom — it is just
unclear who reaches critical mass first.
The transferable rule: when a new interaction medium appears, do not rush to build
apps — check whether the distribution slot is still empty. While everyone builds apps,
the shovel positions (stores, templates, runtimes) face less competition. VoiceOS binds
generation and distribution together: the lower the generation barrier, the more there is
to distribute, and the two feed each other.
What to watch next
① Whether the count of third-party installable voice apps in the store grows in three
months — that is the ecosystem signal
② Whether VoiceOS's own MAU keeps climbing — the channel's value tracks the host
product's activity
③ Whether the first developer actually makes money from a voice app — someone earning is
what makes an ecosystem real
What you can take from it
Product logic: for platform plays, put creation and distribution in one loop — what
users generate feeds your channel by default, which halves cold-start cost. VoiceOS joins
"say a sentence to create" and "share a link to install" end to end, so the two feed each
other's traffic.
Pricing structure: free channel, paid service — claim the distribution slot free
first, charge later. For early platforms, this trade-off is worth copying: free channel,
paid product, then monetize the channel once the volume shows up.
Verdict
Worth watching. The scarcity of the distribution bet exceeds that of VoiceOS itself —
far fewer competitors are building a voice-app store than building voice dictation — and
YC backing gives it room to experiment. But until the ecosystem arrives it remains an
early wager, and the payoff point is whether third-party developers make money on it.
Check back in three months against the app count in the store.