Microsoft AI: MAI-Voice-2-Flash

Origin

microsoft/mai-voice-2-flash

MAI-Voice-2-Flash is a low-latency text-to-speech model from Microsoft AI for voice agents, assistants, call centers, accessibility, narration, and other interactive applications. It generates expressive 24 kHz mono speech across 15...

Pricing

Input · 18 MA / 1M UTF-8 bytes

Billing follows the provider's own reported cost. The unit prices here are for reference; a model marked "billed at actual usage" has no reliable pre-call price, so your usage page is the record of what you paid.

Call it

Speech is served on POST /v1/audio/speech and returns raw audio bytes, not JSON, so send it straight to a file with -o. response_format defaults to mp3. You are billed at the provider's actual reported cost; the ledger records your input text's UTF-8 byte count as the quantity (a Chinese character is three bytes).

curl https://macdecloud.com/v1/audio/speech \
  -H "Authorization: Bearer dcld-sk-..." \
  -H "Content-Type: application/json" \
  -d '{"model":"microsoft/mai-voice-2-flash","input":"Hello from deCloud.","voice":"alloy"}' \
  -o speech.mp3

Python (OpenAI SDK)

The SDK's response object carries the audio bytes; write them straight to a file.

from openai import OpenAI

client = OpenAI(base_url="https://macdecloud.com/v1", api_key="dcld-sk-...")
resp = client.audio.speech.create(model="microsoft/mai-voice-2-flash", voice="alloy", input="Hello from deCloud.")
with open("speech.mp3", "wb") as f:
    f.write(resp.read())