Gemini App Finally Expands To Audio Files
(Tuesday September 09, 2025 @03:00AM (BeauHD)
from the new-and-improved dept.)
- Reference: 0179095856
- News link: https://tech.slashdot.org/story/25/09/09/0030209/gemini-app-finally-expands-to-audio-files
- Source link:
Google rolled out three big Gemini updates: the app [1]now supports audio uploads (with tiered limits for free vs. paid users), Search gains AI Mode in five new languages, and NotebookLM expands to generate reports, study guides, quizzes, and other formats in over 80 languages. The Verge reports:
> According to a Monday [2]post on X by Josh Woodward, vice president of Google Labs and Gemini, audio file compatibility was the "#1 request" to the Gemini app. Free Gemini users max out at 10 minutes of audio, and five free prompts each day. AI Pro or AI Ultra users, meanwhile, can upload audio up to three hours in length. All Gemini prompts accommodate up to 10 files across various file formats, including within ZIP files.
>
> Additionally, Google Search's AI Mode has rolled out five new language options: Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese, thanks to the integration of Gemini 2.5 with Search, according to a company [3]blog : "With this expansion, more people can now use AI Mode to ask complex questions in their preferred language, while exploring the web more deeply." The Gemini-powered NotebookLM software is also getting [4]an update in the form of new report styles in over 80 languages based on a user's uploaded documents, files, and other media.
[1] https://www.theverge.com/ai-artificial-intelligence/774008/gemini-audio-new-languages-notebooklm-reports
[2] https://x.com/joshwoodward/status/1965057589718499756
[3] https://blog.google/products/search/ai-mode-expands-more-languages/
[4] https://x.com/NotebookLM/status/1965106170152013888
> According to a Monday [2]post on X by Josh Woodward, vice president of Google Labs and Gemini, audio file compatibility was the "#1 request" to the Gemini app. Free Gemini users max out at 10 minutes of audio, and five free prompts each day. AI Pro or AI Ultra users, meanwhile, can upload audio up to three hours in length. All Gemini prompts accommodate up to 10 files across various file formats, including within ZIP files.
>
> Additionally, Google Search's AI Mode has rolled out five new language options: Hindi, Indonesian, Japanese, Korean, and Brazilian Portuguese, thanks to the integration of Gemini 2.5 with Search, according to a company [3]blog : "With this expansion, more people can now use AI Mode to ask complex questions in their preferred language, while exploring the web more deeply." The Gemini-powered NotebookLM software is also getting [4]an update in the form of new report styles in over 80 languages based on a user's uploaded documents, files, and other media.
[1] https://www.theverge.com/ai-artificial-intelligence/774008/gemini-audio-new-languages-notebooklm-reports
[2] https://x.com/joshwoodward/status/1965057589718499756
[3] https://blog.google/products/search/ai-mode-expands-more-languages/
[4] https://x.com/NotebookLM/status/1965106170152013888
Slow news day? (Score:1)
Google, really, just an X announcement? Why didn't you write on your own blog platform or use youtube to host the video - if you did, I can't find it.
Also..neat I guess, I like notebooklm for condensing things into an easy to read mindmap that links back to the source material, but this doesn't seem like 'front page newsworthy' enhancements.
Last 'rant', gemini is a great LLM but like many google products it feels pretty fragmented, and their API is just as fragmented with both gemini api and
Re: (Score:3)
And boy is gemini 2.5 flash not good at all compared to the 'free' tier of other providers.
Nor is the free tier of Copilot. My theory is that MS and Google have such big user bases, that they have to be extra stingy with the amount of compute per user, otherwise the cost might just get out of control. Especially with Google doing the AI Overview thing even when people are just trying to do a regular search, and MS similarly inserting Copilot into workflows unasked..