Gladia's latest product offers real-time AI audio processing and transcription
Gladia, a voice AI company, has launched Gladia Real-Time, a groundbreaking low-latency audio transcription and analysis product supporting over 100 languages. Gladia has also secured $16 million in Series A funding to develop its audio AI infrastructure.
In a recent blog post, Gladia recently shared news of the launch of its new product, Gladia Real-Time, and the closing of a $16 million Series A funding round. The Gladia Real-Time transcription API delivers accuracy and speed, transcribing over 100 languages with a latency of under 300 milliseconds and providing both a transcript and AI-generated sentiment analysis, named entity recognition, and summarization insights for a meeting or call in less than a second. Gladia's speed, accuracy, and affordability make it an ideal choice for automated meeting recorders, note-taking assistants, and contact center platforms.
Gladia's journey began in 2022 with a mission to deliver a scalable speech-to-text API that was fast, accurate, and affordable. Further work on its product led to Whisper-Zero, an ASR system designed to reduce hallucinations and fine-tuned on accented voices and noisy backgrounds to deliver exceptional performance even in less-than-ideal circumstances. Gladia's Whisper-Zero system boasts an impressively low hallucination rate and has been adopted by customers including VEED, Livestorm, Method, Recall, and Circleback.
Despite Whisper-Zero's success, a final challenge remained: real-time transcription often sacrifices accuracy in exchange for a shorter wait time while asynchronous transcription delivers accuracy at the expense of speed. With its Real-Time API, Gladia expects its customers will no longer need to compromise one for the other. Moreover, with its recently closed $16 million Series A—led by XAnge with participation from investors Illuminate Financial, XTX Ventures, Athletico Ventures, Gaingels, Mana Ventures, Motier Ventures, Roosh Ventures, and Soma Capital—Gladia plans to continue building and end-to-end audio AI infrastructure.
