Skip to content
Speech to Text

Every word,
captured perfectly

Industry-leading speech recognition. 100+ languages, real-time streaming, speaker diarization — all through one API.

0.5%
Accuracy
0+
Languages
<0ms
Latency
0M+
Hours Processed

What makes it different

Three capabilities that set Nur STT apart.

01

Real-time streaming transcription

Stream audio and receive transcriptions word-by-word in under 500ms. Built for live meetings, broadcasts, and phone calls — no waiting for processing to finish.

LIVE
00:42.37148kHz · 16-bit · PCM
Features

Built for production

Enterprise-grade reliability, security, and scale — ready when you are.

Custom vocabulary

Add brand names, acronyms, and domain-specific terms.

Smart formatting

Auto-punctuation, capitalization, and paragraph breaks.

Word timestamps

Millisecond-accurate timing on every word.

Any format

MP3, WAV, FLAC, OGG, WebM, and 20+ more.

Noise reduction

Transcribe clearly in noisy environments.

Batch processing

Process thousands of files in parallel.

Integrate in minutes

Three lines of code. That's it.

API reference
1from nur import NurClient
2
3client = NurClient()
4
5transcript = client.stt.transcribe(
6 file="meeting.mp3",
7 language="en",
8 speaker_diarization=True,
9 timestamps=True
10)
11
12for segment in transcript.segments:
13 print(f"[{segment.speaker}] {segment.text}")

Start transcribing today

Free tier includes 10 hours per month. No credit card required.