Velma Transcribe
Features
- Transcribes messy real-world audio with overlapping speakers, accents and background noise
- Detects 20+ emotions, identifies accents and labels speakers
- Captures tone, emotion and intent using an ensemble of specialized models
- Streams transcription with sub-second latency and partial transcripts
- Offers batch and streaming transcription through an API
- Stores no audio or transcript data on servers
Pricing Details
💵 Pricing (April 2026 Status) Velma Transcribe uses a transparent, usage-based model designed for high-throughput developers: Plan Price (Approx.) Key Features Free Credits $0 Up to 1,500 hours of free credits for new developers to test the API. Standard API $0.03 / hr Full access to batch and streaming; Emotion & accent detection. Enterprise Custom High-volume discounts; Dedicated support; HIPAA/SOC2 compliance. 🎁 Free Version: Modulate is currently offering a very generous Free Credit program for developers. New accounts can get up to 1,500 hours of transcription for free to integrate and test the ELM architecture in their own apps.