On-screen text
speechma.com
luvvoice.com
AI tools
Unlimited
in multiple
languages
Oh my God
do I sound like
human to you?
reads the Script
naturally
unlimited voiceovers
and it's completely
free?
OpenAI.fm
An interactive demo for developers to try the new text-to-speech model in the OpenAI API.
START BUILDING
VOICE
Alloy
Ash
Ballad
Coral
Echo
Fable
Onyx
Nova
Sage
Shimmer
Verse
VIBE
Calm
Connoisseur
Robot
Cheerleader
Medieval Knight
Sports Coach
Identity: A robot
Affect: Monotone, mechanical, and neutral, reflecting the robotic nature of the customer service agent.
Tone: Efficient, direct, and formal, with a focus on delivering information clearly and without emotion.
Emotion: Neutral and impersonal, with no emotional inflection, as the
the vibe of the voice
Greetings, customer.
Please provide your
Thank you. Verifying
Your return request
returned to our war
payment method with
For any further ass
2. Thank you for sh
Well now, partner,
get you fixed up.
If your internet's
line. Need help wi
it out.
For device setup o
nice and easy. If
talkin' to a real
No need to fret, w
Voice Affect: Calm, grounded, and introspective—like a storyteller guiding the audience
through a meaningful journey.
Tone: Thoughtful, sincere, and observant—friendly but with a sense of quiet authority and
curiosity.
Pacing: Slow and intentional and unhurried, giving the audience space to absorb key
information.
Emotion: Subtle reflection—express a quiet sense of wonder and respect for the
subject matter.
Pronunciation: Clear, with natural emphasis on key terms to highlight
meaning.
Pauses: After major transitions or impactful lines, allowing the listener
to process.
How to Prompt AI Voice Tools Using the "Vibe Format"
Use this structure to guide AI voice generators like OpenAl.fm so your voiceovers sound human,
expressive, and on-brand.
Step-by-step Prompt Guide:
1. Start your prompt like this:
Voice Affect: [Describe the overall sound/mood of the voice. Is it calm, lively, grounded, etc.?]
Tone: [Describe the personality-friendly, professional, empathetic, etc.]
Pacing: [Is it slow and thoughtful? Steady and confident? Fast and upbeat?]
Emotion: [What subtle emotion should the voice express? Wonder, joy, empathy, etc.?]
Pronunciation: [Should the words be crisp? Natural? Emphasized in certain parts?]
Pauses: [Where should it pause for effect, clarity, or emotion?]
Example Prompt:
Voice Affect: Calm, composed, and reassuring-project quiet authority and confidence.
Tone: Sincere, empathetic, and gently authoritative.
Pacing: Steady and moderate-unhurried enough to communicate care.
Emotion: Genuine warmth and understanding.
Pronunciation: Clear and precise, with emphasis on key phrases.
Pauses: Brief pauses between ideas to let thoughts breathe.
You can tweak it for different styles:
For documentaries: "Voice Affect: grounded and reflective..."
For ads: "Voice Affect: upbeat and energetic..."
For tutorials: "Voice Affect: friendly and instructional..."
Identity (Voice Signature):
Identity: [Describe the voice's unique character or match? Is it cloned from a real person? Does it
mimic a gender, age, accent, or celebrity style?]
This controls how much the AI voice sounds like a specific person-useful for:
Voice cloning (you, a client, a narrator)
Character consistency across content
Branded voice personas
Identity Prompt Example (for a cloned voice):
Identity: Matches John's original voice - male, mid-30s, Filipino accent, mellow tone, casual
cadence
Voice Affect: Grounded
Tone: Friendly but confident
Pacing: Steady and conversational
Emotion: Natural optimism
Pronunciation: Clear, Filipino-accented English
Pauses: After important statements, slightly between thoughts
Or You can ask CHATGPT to create the prompt for you.
Ask ChatGPT what style of voice, then paste the prompt guide below. Then just copy the
result, and paste it in OpenAI.fm
I even made
a guide
to help you write
better vibe
instructions
comment
"OPENAI"
and i'll send it to you