BlabbyAI delivers ultra-accurate voice-to-text on Windows and in your browser. Dictate emails, messages, and notes in 100+ languages, faster than you can type.
Free Options
Launch tags:Productivity
Launch Team
Render: Get up to $100K in startup creditsCloud infrastructure for founders who want to move fast and NOT break things
Promoted
Overall, I am genuinely impressed with this software. It is very close to what I had been looking for and fits seamlessly into my daily workflow.
In my work, I frequently use AI tools to structure text, translate between German and English, and convert rough thoughts into properly formatted content. I create many tasks in Microsoft Azure DevOps, and everything there is in English. However, I personally think and draft a lot in German. Constantly switching to ChatGPT, explaining something in German, translating it, formatting it in Markdown, and then pasting it back into Azure DevOps is extremely inefficient. The friction becomes particularly noticeable when dealing with many small text elements, such as Product Backlog Items that require specific Markdown formatting.
What makes this tool outstanding is that it works directly in the browser. In my work environment, I use Windows and cannot freely install software, so a browser extension is the perfect solution. I can stay fully within my context in Azure DevOps, see exactly what I need to see, and dictate my thoughts directly into the active text field. Whisper AI then produces the structured output in the format I want. This context preservation is a major productivity gain.
The custom commands are another strong feature. After a few iterations, I was able to define commands that reliably produce the exact output structure I need. That level of configurability is powerful and significantly reduces repetitive work.
There are, however, a few areas for improvement:
At the beginning, I unintentionally consumed a large number of credits because I selected a model that was unnecessarily powerful for my use case. While this was ultimately my own configuration mistake, a clearer indication of model cost versus typical use cases might help prevent similar situations for new users.
Another limitation is that the extension does not activate on certain websites. This may be due to technical restrictions on those pages or how input fields are defined. Still, it would be highly valuable to have a “force activation” option. For example, if I am focused on a specific text field, I would like to manually trigger the tool and ensure that my dictated input is inserted exactly there.
I also frequently switch between two preferred modes. While the current solution works, changing modes requires multiple clicks. Since the language selection has a quick-access button—even though I rarely change the language—it would be more efficient to have a similar shortcut for switching between commonly used modes. A faster toggle between favorite modes would noticeably improve usability.
Despite these minor points, the overall experience is excellent. The tool does exactly what I was searching for: it removes friction from multilingual, structured knowledge work inside browser-based tools.
One final observation: I was surprised that ChatGPT itself never suggested this solution to me, while Claude Code did. Given how well this integrates into AI-heavy workflows, it feels like a tool that more professionals should be aware of.
In summary, this is a highly practical and well-designed productivity enhancer, especially for users who work extensively with text that has to be in a specific format or style.
@gregor_jakobi Thank you so much, Gregor. Those seem like minor points, but in the long run it can be frustrating. I’ll try to make it even more frictionless!
I’ve been using this product for about two and a half months now, and it’s seriously boosted my average productivity at the computer. Since I have to write a ton of text every day, working with neural networks and all, this thing is saving me at least ten to fifteen hours a month, bare minimum.
What really gets me is this feature where you can create your own mods. So first, you get the standard transcription through the GPT models, but then you can set up your own prompt and process all your recordings exactly how you want. Honestly, I think that’s a killer feature. Before I started using this product, I tried out every similar tool on the market right now, and it was this killer feature that made me pick this one.
For example, I’ve set up mods for myself like being able to speak in my native Russian and instantly get a polished English translation thanks to GPT. Or, say, I can ramble on with messy, unstructured speech and then run it through a complex prompt that turns it into something more logical and organized.
@sergeirom You described it better than I did😅. Thank you Sergei!
I've checked out several similar tools, and while they all get the job done, this one gives you way more flexibility to customize it just how you like.
The two step processing with custom mode is great. I love it.
Finally a tool that actually works!