I've been using Google Colab for Whisper for a long time, but it wasn't always comfortable and efficient. This service allows transcribing at a much lower cost compared to using the OpenAI API or Colab Pro and offers the added benefit of a distributed environment feature. I believe this service will be valuable for numerous cases.
The team is very proactive and responsive. Just today, they published an article on how to use the API in Zapier, which helped me implement Webhooks in Activepieces. This level of support and documentation is greatly appreciated.
From what I understand, privacy considerations are treated with high importance here, which is crucial. However, I would appreciate more detailed information about the specific privacy measures in place.
The speed and accuracy of transcriptions have been impressive so far. The ability to integrate with various automation platforms (Activepieces, Make, n8n, etc) through well-described API and Webhooks adds significant flexibility to the service.
While my experience has been largely positive, I'm curious about potential improvements in the pipeline, such as supporting fast-Whisper and translation.
Overall, this service offers a compelling alternative to more expensive options in the market, especially for those needing reliable, cost-effective voice recognition capabilities. I look forward to seeing how it evolves and improves over time.