MAI-Voice-2 Flash API
MAI-Voice-2 Flash text to speech — low-latency speech for voice agents, billed on input length.
The low-latency tier of MAI-Voice-2, built for voice agents, call centres, and accessibility tooling where the user is waiting on the audio. Same 15-language coverage at roughly a third less per character.
Takes 30 seconds.
MAI-Voice-2 Flash API pricing
Pay per request from your YepAPI balance — no subscription, no monthly minimum. Prices updated October 23, 2026.
| Billing | Price |
|---|---|
| Rate | $0.0317/1K chars |
| What it costs | Per job |
|---|---|
| Short reply200 characters | $0.0063 |
| Blog post narration10,000 characters (~1,600 words) | $0.317 |
| Audiobook chapter100,000 characters | $3.17 |
4,900+
Developers
2.3M+
API calls served
184
Endpoints
$0.01
Per call
Yep, that's it.
Playground coming soon — use the API endpoint directly with your API key.
MAI-Voice-2 Flash specs
| Provider | Microsoft |
|---|---|
| Model ID | microsoft/mai-voice-2-flash |
| Endpoint | /v1/media/queue |
| Type | Text to Speech |
| Output formats | mp3 |
| Voices | 4 presets (default en-US-Harper:MAI-Voice-2) |
How to get a MAI-Voice-2 Flash API key
- 1
Create an account
Sign up at yepapi.com with Google or email. It takes about 30 seconds.
- 2
Create an API key
Open the dashboard, go to API Keys and create a key. Copy it — it starts with yep_sk_.
- 3
Call MAI-Voice-2 Flash
Send a request to /v1/media/queue with the x-api-key header and model "microsoft/mai-voice-2-flash".
- 4
Pay as you go
Usage is billed per request from your balance. Top up any amount; there is no subscription.
Strengths
Tuned for interactive voice where a caller is waiting — sub-second-class synthesis for short utterances.
Keeps MAI-Voice-2's 15 languages and 18 locales, so a Flash-based agent is not linguistically narrower.
Roughly a third less than the full model, which matters when every turn of a conversation is synthesised.
Clean telephony- and app-ready audio without post-processing.
Quick start
Copy this snippet and start making calls with MAI-Voice-2 Flash.
// Step 1: Submit job
const res = await fetch('https://api.yepapi.com/v1/media/queue', {
method: 'POST',
headers: {
'x-api-key': 'YOUR_API_KEY',
'Content-Type': 'application/json',
},
body: JSON.stringify({
"model": "microsoft/mai-voice-2-flash",
"prompt": "Thanks for calling. I can help with that — can you confirm the last four digits of your account number?",
"options": {
"voice": "en-US-Harper:MAI-Voice-2"
}
}),
});
const { data } = await res.json();
const jobId = data.jobId;
// Step 2: Poll for result
const status = await fetch(`https://api.yepapi.com/v1/media/status/${jobId}`, {
headers: { 'x-api-key': 'YOUR_API_KEY' },
});
const { data: job } = await status.json();
// job.status: "pending" | "processing" | "completed" | "failed"
// job.result: { text?, image?, audio?, video? }MAI-Voice-2 Flash vs similar models
Every model below runs on the same API key — switch by changing the model ID.
| Model | Provider | Price |
|---|---|---|
| MAI-Voice-2 Flash | Microsoft | $0.0317/1K chars |
| MAI-Voice-2.1 Flash | Microsoft | $0.0316/1K chars |
| MAI-Voice-2.1 | Microsoft | $0.0464/1K chars |
| MAI-Voice-2 | Microsoft | $0.0465/1K chars |
| Kokoro 82M | hexgrad | $0.0014/1K chars |
| Fish Audio Drama 3 | Fish Audio | $0.0042/1K chars |
| HeyGen Voice Turbo | HeyGen | $0.0084/1K chars |
| Sesame CSM 1B | Sesame | $0.0148/1K chars |
Why use MAI-Voice-2 Flash through YepAPI?
MAI-Voice-2 Flash API — low-latency speech for voice agents
MAI-Voice-2-Flash is the low-latency tier of Microsoft's MAI-Voice-2, aimed at voice agents, call centres, accessibility tooling, and anything else where the caller is waiting on the audio.
On YepAPI it is a single `POST /v1/media/queue` call billed at $0.0317 per 1,000 characters — about a third less than the full MAI-Voice-2 — on the same key that reaches every other speech, image, and video model.
What developers say
“Switched from SerpAPI and cut our SERP costs by 80%. Same data quality, way simpler billing.”
“One API key for AI models, SERP data, and web scraping. Saved us from managing 4 separate providers.”
“The $5 free credit let us prototype our entire rank tracking feature before committing. No other API does that.”
Frequently asked questions
The low-latency tier of MAI-Voice-2, built for voice agents, call centres, and accessibility tooling where the user is waiting on the audio. Same 15-language coverage at roughly a third less per character.
MAI-Voice-2 Flash costs $0.0317/1K chars through YepAPI, billed per request with no monthly fee. For example, a short reply (200 characters) costs about $0.0063.
Sign up at yepapi.com (Google or email, about 30 seconds), open the dashboard and create an API key. The same key works for MAI-Voice-2 Flash and every other model — no separate Microsoft account or waitlist.
Use "microsoft/mai-voice-2-flash" as the model on /v1/media/queue. The short alias "mai-voice-flash" also works.
Explore more models
MAI-Voice-2.1 Flash
MicrosoftMAI-Voice-2.1 Flash text to speech — low-latency expressive speech in 23 languages, billed on input length.
MAI-Voice-2
MicrosoftMAI-Voice-2 text to speech — expressive speech in 15 languages, billed on input length.
MAI-Voice-2.1
MicrosoftMAI-Voice-2.1 text to speech — studio-grade speech in 23 languages, billed on input length.
Orpheus 3B
Canopy LabsOrpheus 3B text to speech — natural English prosody, billed on input length.
Voxtral Mini TTS
MistralVoxtral Mini TTS text to speech — 30 emotion-tagged voices, billed on input length.
Eleven v4 Turbo
ElevenLabsEleven v4 Turbo text to speech — Eleven v4's expressive delivery, tuned for speed, billed on input length.
More Text to Speech models
- Voxtral TTS 3 API · Mistral
- Eleven Flash v2.5 API · ElevenLabs
- Gemini 3.1 Flash TTS API · Google
- Fish Audio S2 Pro API · Fish Audio
- Gemini 3.8 Flash Lite TTS API · Google
- Eleven Multilingual v2 API · ElevenLabs
- Kokoro 82M API · hexgrad
- Fish Audio S1 API · Fish Audio
- Sesame CSM 1B API · Sesame
- HeyGen Voice Turbo API · HeyGen
- Qwen-Audio 3.0 TTS Plus API · Qwen
- Eleven v3 Conversational API · ElevenLabs