Case Study: Wistia achieves more accurate video captions with Deepgram

Wistia improves transcription accuracy by 30% with Deepgram

Wistia, a video marketing platform, needed a highly accurate automatic speech recognition solution to provide reliable captions for its customers' videos. This was essential for improving accessibility and viewer engagement. To address this challenge, Wistia chose to integrate Deepgram's Nova-2 speech-to-text API through the AWS Marketplace.

By implementing Deepgram's solution, Wistia achieved a 30% lower word error rate and up to 40x faster inference times compared to competitors. Deepgram also provided a 3-5x reduction in cost. This integration allowed Wistia to seamlessly offer high-quality transcription features, including speaker diarization and multilingual support, which enhanced the overall value of its platform for its users.

Wistia