Lesan AI provides API services for speech recognition (ASR) and machine translation (MT) optimized for African languages including Amharic, Tigrinya, and English. Use our APIs to transcribe audio, translate text, and build multilingual applications.

Services

ASR (Automatic Speech Recognition)

Convert audio to text with high accuracy. Supports file uploads (sync and async), batch processing, and real-time WebSocket streaming. Speaker diarization and timestamped segments are included in results.

• ASR Guide — File uploads, async polling, batch processing
• Streaming Guide — Real-time WebSocket transcription

MT (Machine Translation)

Translate text between English and African languages using neural machine translation models. Supports bidirectional translation between all supported language pairs.

• MT Guide — Translation API usage and language pairs

Supported Languages

• Amharic (am) — ASR + MT
• Tigrinya (ti) — ASR + MT
• English (en) — ASR + MT

Key Features

• Real-time streaming — Transcribe audio as it is recorded via WebSocket
• Async processing — Queue long audio files and get notified via webhooks
• Batch processing — Transcribe multiple files in a single request
• Speaker diarization — Identify different speakers in audio
• Webhooks — Receive notifications when async jobs complete

Getting Started

• Quick Start — Make your first API call in minutes
• Authentication — Set up API keys and permissions
• Best Practices — Production recommendations
• API Reference — Interactive API documentation