Speaker 1: Which is the best transcription API for your no-code project and what I mean by transcription is I mean speech-to-text or speech recognition. You put in an audio file or a video file and you get text out the other side. I've got first-hand experience using the services I'm about to review building no-code apps with Bubble.io and if you're watching this video it's because you've got an idea and you want to bring it to life using no-code and if you want to accelerate that process to click the link down in the description to check out our website PlanetNoCode.com where we've got hundreds of Bubble tutorial videos, hours of content for you. Let's dive in with the first one though which is Whisper by OpenAI and the first thing to point out of course is that Whisper is a model and if you access it through the OpenAI API, you're just accessing that model. You can access Whisper through other providers, other API providers. The main downside of Whisper is that it has a 25 megabyte upload limit which simply isn't going to cut it if you're transcribing an hour-long meeting or that's just audio. But if you're going to feed video in and you want the audio from that video, it's not going to work with Whisper. You're going to hit that limit very quickly. So this is the next service is Assembly AI and I've been using Assembly AI for years and they are basically, they've got their own model, but it's kind of Whisper with tons of extra features such as speaker recognition, paragraphs, smart formatting, so that you get back something that looks more like a professionally typed transcript. And because they're doing all of the extra processing and it can deal with bigger files, there is an extra step which is to use a webhook. So that means that when you supply Assembly AI with your audio file, you are then giving them a webhook, an endpoint on your bubble app where they can inform you to say that the transcript is done and then you go back and you retrieve the transcript based on a matching ID. So there's a few extra steps. We've got a mini-series looking at Assembly AI if you want to use them and I've been rating and recommending them for years. Maybe I'm a little bit late to the party with DeepGram, but I think it has now become my transcription API of choice. I've noticed DeepGram in particular for its speed in use in services where you've basically got speech-to-speech AI. We've got a video looking at VAPI and DeepGram is one of the providers you can pick there. And of course low latency is important in speech-to-speech AIs. Now this is the reason why I love DeepGram and why I think I'm going to be using DeepGram in a project that I'm working on right now, which is if I hop into the playground, I've uploaded the audio file from one of our recent videos. It's about 16 minutes long. I've added in smart formatting, we'll add in punctuation, paragraphs. These are all features, of course, that are in Assembly AI and then I'm going to hit run and just watch how quick this is. There, it's done. So you can use webhooks with DeepGram, but it may be advantageous to you because you don't have to use webhooks. You can just wait and as the wait isn't that long, in fact, there we go, processing time, five seconds to do 14 minutes of content, you don't have to worry about webhooks. It cuts down the complexity of the bubble app that you're building. So what are your thoughts? Have I missed out a decent speech-to-text service that you want me to check out? Please leave a comment down below.