AssemblyAI provides a platform for building AI applications using voice data. Through its secure and scalable API, developers can access advanced AI models for speech recognition, automatic transcription, and speech summarization. The platform boasts human-level accuracy with its latest AI model, Conformer-2, achieving state-of-the-art results, especially with a significant reduction in errors on noisy data. The API is designed for real-world applications, offering features like speaker labels, word-level timestamps, and profanity filtering. It enables the creation of smarter apps by summarizing, diarizing, detecting sentiment, and more. Additionally, AssemblyAI introduces a new framework called LeMUR for building LLM-powered apps on voice data. It processes terabytes of audio data daily, ensuring reliability and compliance with SOC 2 Type 2 standards. The platform is trusted by various companies and invites developers to explore its AI models and features to enhance audio and video intelligence in their applications.
Key Points about AssemblyAI
- Advanced AI Models: AssemblyAI provides advanced AI models for speech recognition, automatic transcription, and speech summarization through a secure and scalable API.
- Human-Level Accuracy: The platform’s latest AI model, Conformer-2, achieves state-of-the-art accuracy, reducing errors on noisy data significantly.
- Real-World Application Design: The API includes critical features like speaker labels, word-level timestamps, and profanity filtering for practical applications.
- Smart App Building: AssemblyAI enables developers to build smarter apps by summarizing, diarizing, detecting sentiment, moderating content, redacting PII, and more.
- LeMUR Framework: The introduction of the LeMUR framework facilitates the building of LLM-powered apps on voice data.
- Enterprise-Scale Processing: The platform processes terabytes of audio data every day with over 99.9% uptime and is compliant with SOC 2 Type 2 standards.
- Developer-Centric: AssemblyAI is built for developers, offering features like real-time audio transcription, auto-generating subtitles, and identifying speakers among others.
- Customer Success Stories: Companies like CallRail, Grain, and Aloware have successfully utilized AssemblyAI to enhance their services and generate valuable insights.
- Community of Developers: Over 90,000 developers are engaged in building with AssemblyAI, indicating a robust community and trust in the platform.
- Ease of Getting Started: Developers can easily start building with AssemblyAI by trying the API for free, with simple code snippets provided for a quick start.
Pricing
| Service Type | Pricing | Features |
|---|---|---|
| Core Transcription | $0.650016 per hour | – Near human-level accuracy speech-to-text transcription |
| – Dual channel transcription | ||
| – Speaker diarization | ||
| – Export SRT or VTT caption files | ||
| – Auto punctuation and casing | ||
| – Filler word filtering | ||
| – And more… | ||
| Real-time Transcription | $0.75024 per hour | – High accuracy, low latency transcription |
| – Custom vocabulary | ||
| – Auto punctuation and casing | ||
| Audio Intelligence | Various pricing per hour (e.g., $0.30/hr for Auto Chapters) | – Content Moderation, Entity Detection, PII Redaction, etc. |
| LeMUR | $0.017 – $0.049 / 1K tokens (input-output) | – Question & Answer, Action Items, Custom Summary, Custom Task |
| Enterprise | Custom Pricing | – AI at scale with additional concurrency |
| – Custom integrations with an AssemblyAI engineer | ||
| – Dedicated support |
