67 AssemblyAI - Speech to Text API Reviews
I love how AssemblyAI delivers outstanding transcription accuracy even on noisy or low-quality audio. The SDKs, documentation and code samples made integration into our codebase very easy and almost instantaneous. On top of all that, features like custom vocabulary tuning, topic detection, and sentiment analysis mean I can rely on a single platform for everything from basic transcripts to deep audio insights. Review collected by and hosted on G2.com.
Occasionally the API struggles with heavy accents or extremely fast speech, leading to minor mis-transcriptions that require manual correction Review collected by and hosted on G2.com.
...it is a very comprehensive tool that allows not only for a very accurate transcription of the text, but also includes punctuation. It also has various features such as speaker recognition. Very satisfied. Review collected by and hosted on G2.com.
I have nothing to report that I don't like at the moment. Review collected by and hosted on G2.com.
I like the fact that it works in multiple languages, and then you can use speech dialization, and that it's pretty easy to set up in Python. on, I think, basically, the comprehensiveness combined with the simplicity of that API makes it a really great tool. Review collected by and hosted on G2.com.
I think it's more that the tool is so simple that on the back end, There's a lot of things that you have to do as a user, perhaps, to get it exactly how how you want it. I think it would be good if there was more testing in the playground so that it didn't costs loads of money to better understand how to format things or to get things in a way that you most like how to use them. Review collected by and hosted on G2.com.
The documentation is fairly clear, and there's a lot of good libraries to get you going. If you're using any tech stack, or just want to roll your own connection, it's pretty clear what you need to do. Of the several different integrations I'm working with at the moment, this one is the easiest. Review collected by and hosted on G2.com.
Getting to the API documentation is actually not as easy as I'd like. Because everything seems to run so smoothly, I don't have to work with AAI too often, it just does the job. So when I do need to alter something in my usage, I do need to work out how to get to the API documentation. Minor issue, but it something I think about every time I use it! Review collected by and hosted on G2.com.
There were 2 things we were looking for when evaluating speech to text APIs:
1) Quality of the speech to text (the most important thing)
2) Speed of the speech to text
3) Documentation / ease of use.
And after using this service for a few months now, AssemblyAI knocks all 3 out of the park. Review collected by and hosted on G2.com.
The one thing we are having slight issues with is speaker diarization with more than 10 speakers. Honestly, it's a bit of an edge case, but something it doesn't do great with all the time. Anything under that though hasn't been an issue, and that's the vast majority of our use case Review collected by and hosted on G2.com.
It works well and is pretty accurate in usage. Using LeMUR summaries is great and works well but needs a defined prompt based on the use case. Can return multiple formats which is great. LeMUR responses return fairly quick and are easy to set up. Review collected by and hosted on G2.com.
Takes a moment to start up and sometimes punctuation could be better, it often breaks sentences up into smaller bits.
Setting up websockets for streaming is quite difficult due to many parts. Review collected by and hosted on G2.com.
The ability to identify speakers and get a detailed time stamp based division. Review collected by and hosted on G2.com.
The segmentation of dialogues is a little rusty with associating the right speaker. Review collected by and hosted on G2.com.
AssemblyAI speech to text API is really easy to use; I’m not a tech profile and I use it both with automation platforms (such as Zapier) and custom code.
It is cheap, for some use cases it cost almost anything! (For example: understand voicemail).
And, with latest model it is very accurate Review collected by and hosted on G2.com.
It will be perfect if in Zapier I can chose EU residency. Review collected by and hosted on G2.com.
Extremely easy to implement the api into any code Review collected by and hosted on G2.com.
Honestly I haven't had many issues with Assembly AI, potentially the underlying model itself could be a little more intelligent (i.e. it still has errors/incorrect words) However for the speed and cost it is extremely impressive. Review collected by and hosted on G2.com.
Its easy to use
It is cheap. We can transcribe 500 audio in a day for under a dollar
It is reliable. The transcriptions are very accurate
It is fast. Takes very little time to transcribe Review collected by and hosted on G2.com.
It does not support many models. If it supported models like whisper, it would be amazing Review collected by and hosted on G2.com.
