TenLabs
TenLabs.ai
Creative Platform
REALTIME SPEECH TO TEXT

Transcribe live speech instantly.

Scribe v2 is the most accurate Speech to Text model. Scribe v2 Realtime sets the benchmark for live transcriptions — powering agents and real-time applications. Both available via API.
SCRIBE V2 REALTIME

Real-time Speech to Text in under
150 ms with Scribe v2 Realtime

Scribe v2 Realtime uses a streaming-first architecture to turn live speech to text instantly, across 90+ languages.
Live
I'm happy to help. What's your email address?
it's john.doe@me.com
Transcribe live speech
Scribe v2 Realtime captures live speech in under 150 ms with exceptional accuracy — built for agents, meetings, and AI Apps that demand instant understanding.
Accuracy chart
High accuracy and ultra-low latency
Scribe v2 Realtime delivers industry-leading accuracy with sub-150 ms latency, setting a new benchmark for real-time speech recognition.
Voice Activity Detection
Automatically detect when speech starts and stops, segmenting speech with precision for smoother live processing.
Transcribe in 90+ languages
Delivering exceptional accuracy across accents, dialects, and recording conditions.
Live in the API
Build Scribe v2 Realtime into your products with the API, with full-streaming support and commit control.

AI Speech to Text transcription across 90+ languages

Our AI speech to text transcription supports 90+ languages, just select the language and upload your audio file.

Built for every workflow,
from API to agents

Speech to Text APIs and SDKs
Integrate Scribe v2 and Scribe v2 Realtime into your product with the API or SDKs.
Speech to Text APIs and SDKs
TenLabs Agents
Enable real-time voice interactions with instant, low-latency transcription.
TenLabs Agents
TenLabs Studio
Convert recordings into editable text, captions, and repurposable content.
TenLabs Studio

Frequently asked questions