AssemblyAI vs noScribe

AssemblyAI provides Voice AI models and APIs for speech-to-text, speech understanding, and voice applications. Its Speech Understanding API can identify and diarize speakers, summarize speech, extract action items and key phrases, detect sentiment, entities, and topics, and handle formatting, language detection, translation, PII redaction, and content moderation. The platform lists support for 99 languages and translation into 86. It offers streaming and voice-agent latency below 300ms and states that it processes more than 800 million API calls per month. Use its managed cloud or self-host models in your own environment; the service also lists API access and integrations including Twilio, Zoom RTMS, Zapier, and LangChain. Security options include TLS 1.2+ in transit, AES-256 at rest, zero data retention, and opting out of model training. A free tier is available. Pay-as-you-go billing is based on usage, with monthly invoices and no minimum commitments. Universal-2 is listed at $0.15 per hour of audio; Universal-3.6 Pro Realtime is $0.45 per hour.

noScribe is a free, open-source desktop app for transcribing interviews and sensitive recordings with AI models running locally on the computer. It uses Whisper-based models for transcription and pyannote for speaker recognition, with support for around 60 languages. Its editor aligns text with audio playback so users can review and correct transcripts. Research tools let users mark pauses and overlapping speech, add timestamps, and choose whether filler words are included. The app accepts almost any audio or video format and exports formatted HTML, plain text, or WebVTT subtitles. Transcripts can be opened in MS Word or LibreOffice, or imported into QDA software such as MAXQDA or QualCoder. The maker estimates a one-hour interview takes about one to three hours to transcribe; NVIDIA CUDA acceleration can reduce processing time. Current versions support Windows 10 and require macOS 14 or later. Linux binaries are not available for version 0.8; users are directed to version 0.7 or manual installation. It is licensed under GPL-3.0.