What I like most about Whisper is how quickly it converts recorded speech into text. I’ve used it during my work with Alignerr to process audio, and it saves a lot of time compared with manually transcribing recordings. The support for different languages and accents is also useful when working with varied audio. Review collected by and hosted on G2.com.

The main downside for me is that setting up and fine-tuning Whisper can take some effort depending on the use case. It also doesn't provide much of a ready-made user interface, so you often need to build the surrounding workflow yourself. Better out-of-the-box tooling and simpler configuration would make it easier to use. Review collected by and hosted on G2.com.

The best part is that it's completely open-source and I can run it locally in Python without needing any paid API keys. I used it for adding speech-to-text features in my college project, and it handles different accents and slight background noise surprisingly well. Even the smaller models like 'base' do a solid job for basic audio transcription. Review collected by and hosted on G2.com.

If you don't have a laptop with a decent dedicated GPU, running the larger models like 'medium' or 'large' can be pretty slow on the CPU. On a normal laptop, you're mostly limited to using 'tiny' or 'base' models, which are fast but can occasionally mishear a few words. Review collected by and hosted on G2.com.

I love how accurate OpenAI Whisper is. Even with accents or low voices, it picks up everything extremely well, actually hearing better than a human would. It's very accurate in all languages and handles people talking in many languages at once, which is amazing for me since I juggle a lot of languages and am bilingual. It is the best solution out there for transcription, much better than Google's Gemini transcription. I also appreciate that it's cost-effective for various use cases, like transcribing content or when I need an intelligent brain dump for productivity apps. The initial setup was extremely smooth in Python notebooks, and I didn't even need to add APIs, making it the easiest setup for me. Review collected by and hosted on G2.com.

I think it's harder to set up inside of apps; I usually set it up myself in Python Colab notebooks, which is easy, but I don't know how to set it up on my website. It doesn't have very good or easy documentation, and even OpenAI's chat model struggles to help understand it. It's too complicated at the moment, and I wish they had a plugin or something that sets it up for you. Pricing was also an issue, leading us to switch to Gemini transcription for their cheaper option and free tier. Review collected by and hosted on G2.com.

The best thing I like about this platform is its versatility and a natural level conversational experience which is very human. We can also use it for research, brainstorming, and other content creation, along with data analysis, coding and summarising documents within the same ecosystem. Right from file uploads, web search, image generation and other projects, this platform makes very useful for both quick questions and other ongoing work periods. It also understands follow-up instructions well so we can refine an output without even restarting the task. Overall, this platform saves us considerable time across many everyday and professional activities. Review collected by and hosted on G2.com.

This platform can occasionally provide inaccurate or overly confident answers and so we need to check all important facts and calculations which are still need verification. The quality of its response can vary depending on how clearly the prompt and context are provided. Their usage limits and access to advanced tool also differs between subscription plan. Longer conversations may also need some time and repetitive, specific instructions to keep the output aligned. I would also appreciate clearer source attribution and more consistency across complex and other multistep task. Review collected by and hosted on G2.com.

What I liked the most was the transcription accuracy. We tested it with different accents and even audio that wasn't very clean, and the results were still really good. I also used it while building a voice bot prototype, and integrating it was quite simple. Overall it just worked well without much tweaking. Review collected by and hosted on G2.com.

The only thing I felt was that it can be a bit slow for real-time use, specially if you're running the larger models yourself. Also, speaker diarization isn't built in, so we had to use another solution for that. Other than that, I didn't face any major issues. Review collected by and hosted on G2.com.

I use OpenAI Whisper for diarization in transcription projects at my localization company. It takes a first pass at transcription work, making the job of the subsequent human easier, which is really helpful. I like that it provides generally grammatically accurate verbatim transcription work. The fact that the output is accurate in single speaker transcriptions is critical to our work, especially for legal transcriptions. I also found the initial setup of OpenAI Whisper to be 'easy peasy.' Review collected by and hosted on G2.com.

The diarization of multi-speaker files has not improved in the last year. The segmenting of different speakers in the transcript is awful and has not improved over time. The software can't even identify when speakers change, even if the changes are very clear and obvious, and I am consistently surprised when it doesn't voice-identify. Review collected by and hosted on G2.com.

What I find most impressive about Whisper is how incredibly helpful it is when you need to extract text from audio. It feels like a superpower for productivity—whether you are dealing with a long meeting recording, an interview, or even just quick voice memos, it handles the transcription with such high accuracy that it saves hours of manual work. Review collected by and hosted on G2.com.

It’s been a lifesaver for turning audio into text, but it can also be frustrating. It often can’t tell who’s speaking, sometimes makes things up, and really needs a powerful computer to run smoothly. Review collected by and hosted on G2.com.

I like that OpenAI Whisper is more hands-free, making it easier for me to speak instead of type when I'm unable to. It picks up my speech well, even when it's a bit unclear, which is super helpful. I appreciate being able to speak my thoughts for assignments without typing them out. Review collected by and hosted on G2.com.

Sometimes it didn’t pick up my speech well or tried to decipher it but it gave me the wrong word. At the time I couldn’t touch my device so I couldn’t stop the speech recognition so it kept listening for my words. Also because I wasn’t talking loud it didn’t get the right words. Review collected by and hosted on G2.com.

OpenAI Whisper delivers highly accurate speech-to-text transcription across multiple languages and accents.

It performs reliably even with background noise and varying audio quality.

Its multilingual support and automatic language detection make it versatile for global use.

Being open-source, it is easy to integrate and customize for a wide range of applications. Review collected by and hosted on G2.com.

OpenAI Whisper can be slower when processing long audio files, especially on lower-powered devices.

Accuracy may decrease with heavy background noise, overlapping speakers, or highly technical terminology.

It occasionally misinterprets uncommon names or industry-s Review collected by and hosted on G2.com.

I like how OpenAI Whisper makes converting long video essays or extended audio into an easier-to-understand document. It makes work easier and less stressful, especially when your job is to sort through tons of podcasts and documentaries. Also, the initial setup was very easy. Review collected by and hosted on G2.com.

I find that sometimes the system will convert the wrong thing into the summary by replacing key information with less useful information. I also noticed that it sometimes gets the translation wrong. Review collected by and hosted on G2.com.