Video Reviews
Think Audio to Text go for the Deepgram
271 Deepgram Reviews
Overall Review Sentiment for Deepgram
Log in to view review sentiment.
Generally, most 3rd party APIs are either just ok or they suck. Deepgram is one of the few along with maybe Stripe that genuinely feel like they wrote their documentation with actual devs in mind, covering everything from use cases to the inner workings of their classes. Very easy to implement and get started with for how complex transcription can be, and it still feels like a cheat code for it to be this fast and affordable. Review collected by and hosted on G2.com.
Nothing, which is rare. My only wish is that it was easier to work with live audio data in a more declarative way. Additionally, I think they need more documentation around system design, e.g. building with Deepgram in a scaleable way with websockets, cloud services, etc. Review collected by and hosted on G2.com.
Since we started using Deepgram, it has been incredible. The best part is its low latency and really fast, accurate transcription. Before Deepgram, we struggled to transcribe videos quickly enough; it often took too long and sometimes failed. I looked at several providers, but Deepgram was the fastest and most accurate. Plus, we didn’t need to host a transcription model, making it much easier for our engineering team to build. Their team has been easy to work with, and we were able to get up and running pretty quickly. We use it frequently for our transcription needs and are planning to experiment with several of their models and their new Text-to-Speech feature. Review collected by and hosted on G2.com.
One thing that would’ve been more helpful is having an endpoint to automatically generate high-quality chapters with timestamps for videos. I would’ve loved a simple endpoint with customization options for this. Ultimately, we built this in-house and are happy with how it came out, but I would’ve preferred to spend more time on other roadmap items. Review collected by and hosted on G2.com.
Deepgram has been an invaluable partner for Insight Health. I evaluated a few other Speech-To-Text products, and Deepgram has the lowest latency, fewer errors, and well-documented APIs. We got HIPAA BAA signed quickly and had everything up and running in production within a week. Deepgram is easier to do business with as a startup. They have a great startup program to get started and scale with them.
They iterate quickly, releasing new models, features, and products—Nova 2 and Aura are examples. If you're an early-stage startup looking for reliable, efficient, and innovative NLP solutions, Deepgram is hard to miss. Review collected by and hosted on G2.com.
One area for improvement is their logging and troubleshooting capabilities. Currently, the logging is somewhat limited, making diagnosing and resolving issues challenging. Enhancing the logging features would greatly aid in troubleshooting during issues. Review collected by and hosted on G2.com.
Deepgram offers impressive speech recognition capabilities backed by powerful deep learning. Its real-time transcription is fast and accurate, and the developer-friendly API simplifies integration. Whether analyzing calls, creating captions, or powering voice bots, Deepgram delivers consistently high performance with flexible customization. Overall, it’s a fantastic tool for anyone seeking reliable speech-to-text solutions. Review collected by and hosted on G2.com.
We would love to see more language support, a custom voice, and even better cost over time. Review collected by and hosted on G2.com.
What I liked most about Deepgram is that it has one of the highest transcription accuracy levels. It is unique because of all the options, such as live captions and subtitles in different languages. This high level of accuracy reduces the amount of time that is required for manual edits and thus more time can be spent on client relation, so it has greatly enhanced the efficiency of our work. Review collected by and hosted on G2.com.
The one drawback of Deepgram is the possible delay that users may experience especially during the periods when many people are browsing the site. This delay can become rather irritating, particularly if we want a real time transcription during a meeting. Furthermore, they could improve the responsiveness of their customer support agents. Sometimes the problems cause delays and to get a solution it may be time consuming. These weaknesses have potential to compromise our system and cause slight hitches which affect the speed at which we reply to our clients. Review collected by and hosted on G2.com.
Deepgram is great for speech to text transcription. Very reliable most of the time. Review collected by and hosted on G2.com.
The enterprise support team takes a long time to get back to you sometimes. Review collected by and hosted on G2.com.
Having tried multiple transcription vendors, one of the reasons we've stuck with Deepgram is that it gives us super quick transcriptions for a decent price. Other vendors we tried were a lot more expensive with higher latency on a 90 minute transcription. Review collected by and hosted on G2.com.
Sometimes the Deepgram algorithm over emphasises custom keywords...replacing previously correct keywords with a custom keyword. This is likely an industry wide challenge as the ASR technology space matures. Review collected by and hosted on G2.com.
Deepgram did a great job building a community around their speech-related AI models. Their support to the community is phenomenal, whether it's about implementation issues or resolving support tickets. We also appreciate that they have a special medical AI model, which has helped us tremendously in our venture. Lastly, their transcription is almost real-time, which is fantastic. We are very excited to try their streaming model as well. The documentation is very comprehensive and very easy to understand. We use it almost for everything related to Speech to Text. Review collected by and hosted on G2.com.
It's a bummer that they don't support Arabic yet. Review collected by and hosted on G2.com.
The transcriptions Deepgram produce are of a very High Quality, not just for English, but for Hindi as well as mixed Hindi + English. Furthermore, Deepgram also has very low latency. Much lower than OAI. Review collected by and hosted on G2.com.
Our developers are still finding it hard to integrate due to lack of documentation. Sometimes I get error in the playground as well. If you create some examples that are colab supporting as well as on js, it will make it much easier to implement it in our workflow. Review collected by and hosted on G2.com.
