Since January 27th, I have been testing Perplexity's new assistant for Android, which was announced on January 23rd, 2025, on Perplexity's official X account.
In a world where AI has gained much relevance since 2024, Perplexity stands out for offering searches with the help of large artificial intelligence models, but what they have presented is truly amazing. I have used Perplexity since 2023; it's a tool that has helped me greatly in my life as a student and worker.
For several Android generations, you can choose the assistant you want, replacing Google's default, which by the way, I have always detested - even Gemini remains somewhat stupid. Perplexity assistant arrives to demonstrate the true power of a real assistant.
The most notable capabilities of Perplexity's voice assistant on Android include:
I have tested each of these functions, and that's why I've decided to share my experience during these 15 days. Let's begin.
Accessing the Perplexity assistant can be done in three ways:
I use navigation gestures on my device, so mainly during these 15 days I've activated it using the power button, the animation shown when activating it is great. Also, if you have a phone with a semi-curved screen, the sensation is very futuristic.
The activation is fast, effective, and has a very well-crafted design, I really like it.
In the interface, you have quick access to activate camera search options, settings, and written search using the keyboard, useful for those moments when you can't speak very loudly, like in a meeting or class for example.
I've really loved the Perplexity icon animation in a sick way, it's simply stylish, futuristic, and minimalist, extra point to that group of designers.
The assistant has good language support, it can automatically detect the system language. Additionally, it's available to choose between English, Czech, French, Standard German (Deutsch), Hindi, Hungarian, Italian, Japanese, Korean, Simplified and traditional Chinese, Polish, Portuguese, Croatian, and Spanish.
It's important to mention that despite your system's defined language, you can go to settings and decide which language you want to communicate with the assistant in, I liked that a lot.
The disadvantages I saw, as a speaker of two languages (Spanish and English), sometimes it caused me a bit of stress having to change the assistant's language so it could understand my requests, I couldn't say "set a reminder" in Spanish while it was configured in English because it wouldn't understand me. I know it would be ideal to keep it in the native language, but it would be useful as a suggestion for this assistant to automatically detect the language you're speaking in, and respond in the same language.
Love this processing animation btw...
For translating text it's good, you can activate the assistant and ask what something says, it will detect what you're seeing on the screen, whether it's a social media post or a photo with text, the assistant will automatically translate it for you.
Also, certain songs or names weren't identified, I tried telling it to play certain K-Pop songs, and it ended up giving me some mismatches in the results, that's something to improve.
Now, referring to the topic of page summaries, there's a small inconvenience, although the summary is useful, and the function of being able to automatically send it through email is incredible...
It only summarizes what's visible on the screen, not the complete information from the website.
In the future, it would be good if the assistant asked the user if the on-screen summary is sufficient, or if they want to summarize the complete article, for this, they would need to share the link with the assistant to get a better summary.