{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"The Good Tech Companies ","title":"Cloud or Local Speech-to-Text? We Built the Same App Both Ways. Here's What We Learnt.","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/08d2cf32\"></iframe>","width":"100%","height":180,"duration":579,"description":"\n        This story was originally published on HackerNoon at: https://hackernoon.com/cloud-or-local-speech-to-text-we-built-the-same-app-both-ways-heres-what-we-learnt.\nHow a hackathon dictation app running Speechmatics on-device speech-to-text exposed why real-time diarization needs GPU acceleration, CoreML, and DirectML.\nCheck more stories related to undefined at: https://hackernoon.com/c/undefined.\n            You can also check exclusive content about #speech-to-text, #local-ai, #on-device-ai, #speech-to-text-workflows, #local-text-to-speech, #local-llm, #ai-model-running-locally, #good-company,  and more.\nThis story was written by: @speechmatics. Learn more about this writer by checking @speechmatics's about page,\n            and for more stories, please visit hackernoon.com.\nA hackathon dictation app runs Speechmatics speech-to-text with zero cloud calls and zero API keys. Real-time diarization still needs a GPU, but not for the reason you'd assume.\n        \n        ","thumbnail_url":"https://img.transistorcdn.com/HZ9CRzf5js9DK86xzUVMWBRbXYwg4dA8xVXJGVzpL6Y/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS8xMTNl/MjgwMmI0ZmEzNThj/YmJiOWNiN2UyZmRm/MzY3My5qcGVn.webp","thumbnail_width":300,"thumbnail_height":300}