Pulse — Speech-to-Text by Smallest AI

Real-time speech-to-text API with sub-100ms latency, diarization, and PII redaction.

🧠 AI Agents

30 views

🎵 Audio & Music

📝 Transcription & Notes

Freemium

About Pulse — Speech-to-Text by Smallest AI

Pulse is Smallest AI's speech-to-text API, built for real-time voice applications like voice agents, live call transcription, and captioning, alongside a high-throughput batch mode for recorded audio. It streams over WebSocket with sub-100ms time-to-first-token at low concurrency, and includes speaker diarization, word- and sentence-level timestamps, keyword boosting for up to 10,000 custom terms, voice-activity events, and automatic inverse text normalization. Pulse supports 21 languages on streaming and 26 on batch, with regional auto-detect and mid-session code-switching. A companion model, Pulse Pro, trades multilingual support for English-only batch transcription tied for #2 on the public Open ASR Leaderboard, built for high-volume call center, meeting, and financial-audio transcription. Both models include built-in PII and PCI redaction (currently English and Hindi), plus optional emotion and gender detection on batch jobs. The platform holds ISO 27001, SOC 2 Type II, HIPAA, and GDPR compliance, with on-premise deployment available on Enterprise plans. Common use cases: live voice agents and conversational AI, call center and contact center analytics, meeting and financial-audio transcription, multilingual customer support, and compliance-sensitive workflows requiring redaction.

Key features

✓

Sub-100ms streaming transcription latency

✓

21+ languages across streaming and batch

✓

Built-in speaker diarization and timestamps

✓

Automatic PII/PCI redaction (English, Hindi)

✓

Pulse Pro: #2 on Open ASR Leaderboard

smallest.ai/speech-to-text?utm_campaign=vizup