Tuned for Persian
Trained and tuned on Persian speech, including everyday spoken Persian. English is supported too.
If you have Persian or English audio and need its exact text with high accuracy, this service is built exactly for that.
Typing out recordings by hand takes hours, and general speech-recognition tools often get Persian wrong: names, numbers, everyday spoken Persian and sentences that mix Persian and English come out garbled.
How it runs
Trained and tuned on Persian speech, including everyday spoken Persian. English is supported too.
Phone calls, meetings, interviews, lectures and voice messages. Send the file, get the text.
Send audio to stt.agentise.ir/v1 from any programming language or tool. OpenAI-compatible.
You pay only for the audio you convert, with a clear usage dashboard.
OpenAI-compatible transcription: send a Persian or English audio file and receive its text.
Early access
Get early accessfrom openai import OpenAI
client = OpenAI(
base_url="https://stt.agentise.ir/v1",
api_key="YOUR_AGENTISE_KEY",
)
client.audio.transcriptions.create(file=open("voice.mp3", "rb"), ...)Email us your full request. Our support team reviews it and calls you back as soon as possible.
Put this in your email
What happens next