AWS has introduced usage-based pricing for bidirectional streaming inference on Amazon SageMaker, addressing a gap in how model providers bill for real-time speech AI services. Previously, providers on AWS Marketplace struggled to charge for consumption during live streams, as the total usage wasn't known until the session ended. This new feature allows providers to report actual usage through a dedicated metadata connection, aligning charges with consumption rather than just endpoint availability. Deepgram is the launch partner, enabling usage-based billing for its speech-to-text and text-to-speech services on AWS Marketplace. AI
IMPACT Enables more accurate billing for real-time AI streaming services, potentially encouraging wider adoption of such applications.
RANK_REASON This is a product update for a cloud service, not a frontier model release or significant industry shift.
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →