Vesper listens, understands and answers out loud, in her own British voice. She’s already built, hosted and maintained. You add one API key and ship her this week.
You never host her, train her or maintain her. Your app posts what your user said; Vesper sends back what to say, and the audio to play.
Choose your monthly volume and create an account. No sales call, no procurement.
It arrives straight away. It’s the only credential your app needs.
One request per turn. Text back, and audio in her voice if you want her to speak.
Every improvement we make reaches your app instantly. You ship nothing.
One request per turn. Text back, and audio in her voice whenever you want her to speak.
const r = await fetch("https://api.meetvesper.co.uk/v1/say", { method: "POST", headers: { Authorization: `Bearer ${process.env.VESPER_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ message: userSaid }), }); if (!r.ok) return fallbackReply(); // carry on without her const { text } = await r.json();
curl https://api.meetvesper.co.uk/v1/say \ -H "Authorization: Bearer $VESPER_KEY" \ -H "Content-Type: application/json" \ -d '{"message": "Anything I have missed?"}' # → {"text": "Not much. A couple of messages from Sarah…"}
{
"text": "Not much. A couple of
messages from Sarah, and your
four o'clock reminder.",
"cached": false
}
vsk_test_ never spendInput · up to 2,000 charactersHer voice, her judgement, her safety net and your controls. All of it built, all of it running.
A stock iPhone ships one British voice, a man at the flat compact grade. Android ships a different woman. Vesper is synthesised centrally, so she sounds identical in every install.
Spending caps by key and by day. When a cap is reached she declines the call before anything is spent.
vsk_live_…a91fvsk_live_…07bcRevoked · 0.2 sShe tells struggling from crisis apart, and answers with the helplines for the user’s own country.
Speaking is what costs, so speaking is what’s metered. Text is bundled generously because it’s nearly free. No seat fees, no minimum term.
A first release or a side project, with room to grow.
One app finding out whether their users talk to her.
A live app with real daily voice use.
An app shipping her to a whole user base.
Dedicated capacity, your own voice, a signed SLA.
The IDE with Vesper built in. Ask her to build, fix or change your project and watch it happen in one window: the conversation, your code, a terminal and a live preview.
Tell her what you want. She reads your code, makes the change, runs it and shows you exactly which files changed.
The conversation, your files, a terminal and a live browser preview, side by side.
The first time it opens, you pick a project. That folder is the only place she may read, change files or run commands.
Download it, install it for yourself with no admin prompt, and start. There is nothing else to install.
The licence in six plain lines. No small print hiding behind them.
A licence to call her for as long as you’re subscribed. She stays ours.
You can’t reverse-engineer her, or resell access on to somebody else.
We update her when we choose to, and your app gets it without shipping.
Rates can change, but never without a month’s warning first.
Provided as-is, and supported properly. Enterprise adds a signed SLA.
We process their words to answer them. Platform and above includes a DPA.
Pick a plan, take your key, and have Vesper answering your users before the end of the day.