The voice of flexVox.
Friendly, direct, practical. A knowledgeable friend explaining a tool they actually use — not a salesperson, not a manual. Below: what we say, what we don't, and how to tell the difference.
Concrete over abstract.
We describe what happens. We don't describe what it enables. Things in flexVox do specific, observable things — they parse, they cast, they generate, they regenerate, they mix, they export. None of them empower anything. We never use the word "leverage" in public copy.
Honest about limitations.
flexVox depends on ElevenLabs for voice generation. We say so on the home page, on the pricing page, and in the press kit. We say so when you open the app. If audio quality matters, it depends on ElevenLabs's catalog and pricing — not on what we wish were true.
flexVox is not a general-purpose audio editor. It does one thing well: turn a multi-speaker script into produced audio. If you need granular waveform editing, beat matching, or multi-track mixing, a desktop DAW is the right tool. We say so on the home page.
Show the workflow, not the technology.
Users care about going from a paste to a finished M4A. They do not
care about REST APIs, SwiftData, the AVAudioEngine tap, or
kAudioOutputUnitProperty_CurrentDevice. We mention
SwiftUI, SwiftData, and Keychain only as proof points of native iOS
construction — not as features. When we do, we say what they mean
to a user: "Your projects persist locally. Your API key never
leaves the Keychain."
Short sentences first.
Lead with the point. Add detail only when it helps. The opening paragraph of any flexVox page should make sense even if you stop reading after the first three sentences.
Words we use.
Paste · assign · generate · mix · export · turn · take · variant · script · speaker · voice · character · review · adjust · regenerate · compare · quick · simple · built-in · native.
Words we avoid.
Seamless · revolutionary · cutting-edge · game-changer · empower · leverage · unlock · supercharge · "AI-powered" as a leading adjective · effortless · best-in-class · world-class · next-generation.
Sample sentences.
- "Paste your script, pick voices, and export a finished podcast — all from your phone."
- "The parser detects speakers and sound effects automatically. You just confirm the ones it's unsure about."
- "If a line doesn't sound right, regenerate it. Your other audio stays untouched."
- "Demo mode lets you walk every screen with no API key. When you're ready for real audio, add the key."
The same idea, three voices.
Below: how flexVox might describe the same feature in three registers. The middle one is the flexVox voice. The others are nearby pitfalls.
"Unleash the power of revolutionary AI voice technology with our game-changing platform — empowering creators to effortlessly transform any script into a stunning multi-voice masterpiece."
"Paste a multi-speaker script. Assign a voice to each character. flexVox generates speech, sound effects, and music — and gives you the tools to regenerate a single line without touching the rest."
"An iOS application that programmatically interfaces with the ElevenLabs API to produce text-to-speech audio outputs, formatted as AAC-encoded M4A files via AVAudioEngine."
A flexVox script, narrated in our voice.
Just because we wrote it. Below: a short scene that demonstrates the relationship between script, expression tags, and SFX. This is the kind of thing flexVox is for.
[SCENE: cold open · 00:00]
ALEX: [warmly] So — first time on the porch.
JORDAN: [a little nervous] First time. I'll try not to spill.
[SFX: distant rain on tin roof (8s) @underlay volume=0.35 loop]
ALEX: [laughs] You wouldn't be the first. Mug to your left.
[Music: warm rhodes hush (10s) @underlay]
JORDAN: [whispers, almost to themselves] It's quieter than I expected.
ALEX: [softly] That's the porch part.