Google Gemini just rolled out a set of new features that genuinely feel like a big step forward. The headline update is Flow Music, a new music creation app that makes it possible to generate beats, vocals, riffs, full songs, mixes, and even music videos from a conversational interface. But that is only part of it. Google also updated AI Studio, added useful sharing customization to NotebookLM, introduced skills in Chrome, and launched a new Gemini desktop app for Mac.
If you use AI for productivity, creative work, or experimentation, these updates matter. Some of them are immediately useful right now. Others are clearly laying the foundation for something much bigger that is coming very soon.
This article breaks down what changed, how each feature works, and where the real value is.
Why These Gemini Updates Matter
There are two big themes running through all of these releases.
- Google is making Gemini more practical for everyday work by turning prompts into reusable tools, improving sharing, and giving users more control over access and limits.
- Google is also expanding Gemini into a broader AI ecosystem with specialized apps and desktop experiences instead of keeping everything trapped in a single chat box.
That is why Flow Music is interesting. It is not just another model update. It is a purpose-built interface for music creation. The same pattern shows up in the desktop app and in Chrome skills. Gemini is becoming less like a single assistant and more like a platform.
Flow Music: Gemini’s New AI Music Creation App
The biggest release here is Flow Music. Google has moved beyond the older, simpler approach of using Lyria inside Gemini and introduced a dedicated music creation experience at flowmusic.app.
The core idea is straightforward: instead of treating music generation like a one-off prompt, Flow Music gives you a working session where you can collaborate with AI the same way you would work with a producer.
What You Can Create in Flow Music
Inside a new session, Flow Music lets you generate a wide range of creative assets, including:
- Music tracks
- Vocals
- Riffs
- Entire songs
- Mixes
- Music videos
- Projects, playlists, and spaces
That alone is pretty wild. But what makes the app more interesting is the amount of input you can feed into the system.
How Input Works in Flow Music
Flow Music is not limited to text prompts. You can also:
- Record live audio directly into a session
- Upload an image to inspire lyrics
- Upload your own audio as source material
That means the tool can support both pure ideation and more hands-on production workflows. You can start from scratch with a prompt, or you can bring in your own voice, concept art, or rough demo.
The “Producer” Layer Is the Real Upgrade
One of the smartest additions in Flow Music is the ability to customize the producer. Instead of generating a song in a vacuum, you can personalize the production system with:
- Specific instructions
- Preferred flows or styles
- Memories and context
This shifts the experience from “generate me a song” to “work with me like a producer who understands my taste.” That is a much more useful model for creators because it makes iteration easier and more consistent.
There are also settings for the underlying components:
- Instrument/model: at the moment, Lyria 3 Pro is the recommended option
- Ghostwriter: available in standard mode, with a Pro option tied to subscription access
- Producer mode: standard for quality and stronger reasoning, or fast for quicker replies
It is also possible to explore playlists, videos, and spaces for inspiration before building something of your own.
A Simple Example Prompt
A good example of how conversational the workflow is comes from a prompt like this:
“I want to create a song and a beat that represents Malibu, California.”
From there, Flow Music builds a response around that concept. In this case, the generated direction was a chill, sun-drenched surf rock sound with smooth bass, shimmering electric guitar, laid-back coastal energy, and a 95 BPM tempo.
Then it generated multiple versions, which is important. Instead of getting one output and hoping it works, you get options. That is closer to an actual creative workflow where you compare, tweak, and refine.
What Happens After the Song Is Generated
Once the music is created, you can keep working on it. Flow Music supports actions like:
- Remixing the result
- Downloading the track
- Sharing it
- Adding vocals
- Requesting edits conversationally
That last part is probably the biggest productivity win. If you want to change the mood, add an element, shift the energy, or rework the arrangement, you just tell the system what you want as if you were speaking to a producer.
There is even a turntable-style feature in the app, which adds to the feeling that this is trying to be a full creative environment rather than a simple generation tool.
Who Flow Music Is Best For
Based on what is available now, Flow Music looks especially useful for:
- Creators who need original background music
- Musicians who want rapid ideation and rough production
- Marketers and content creators experimenting with custom audio and music videos
- Anyone curious about where AI music workflows are heading
It is early, but it already feels like one of the more ambitious Gemini product releases in a while.
AI Studio Now Supports More Flexible Upgrades
The next notable update is inside Google AI Studio. Google has added a clearer upgrade path that gives power users more ways to unlock access.
You now have two main options:
- Add a Gemini API key, which lets you pay per request and experiment directly with Gemini API models and features in the UI.
- Use a Google AI subscription, which gives higher limits and additional Google AI benefits if you already subscribe.
What the API Key Option Unlocks
If you add a Gemini API key in AI Studio, you gain:
- Higher usage limits
- Access to all models
- Access to all agents
- Direct experimentation inside the UI
For serious builders, that matters. If you are testing prompts, models, or workflows at scale, you want the broadest access possible.
What the Google AI Subscription Adds
If you already pay for Google AI, you can use that subscription in AI Studio too. That gives you:
- Higher limits
- Additional Google AI benefits
The main limitation is that this option does not unlock every model and agent in the way the API key route does.
The Best Setup for Power Users
If AI Studio is central to your workflow, the smartest move is to use both:
- A Gemini API key for complete model and agent access
- A Google AI subscription for higher limits and additional perks
This update may seem minor compared with Flow Music, but it points to something important. Google is trying to make AI Studio more serious as a workbench for people who build and experiment regularly.
NotebookLM Now Lets You Customize Notebook Covers and Summaries
NotebookLM also picked up a practical update, especially for people who share notebooks with clients, teams, or collaborators.
You can now customize a notebook with:
- A cover image
- A custom title
- A custom notebook summary
Why This Matters
If you use NotebookLM privately, this may not change much. But if you send notebooks to other people, presentation matters. A polished cover image and a clear summary make the notebook feel more intentional and easier to understand at a glance.
The cover image should be in a 16:9 format, and you can upload your own file. You can also remove or replace it whenever you want.
The custom summary feature is useful because it lets you override the default description with something more specific. That means when someone opens the shared notebook, they get the exact context you want them to have.
An Example Use Case
A notebook titled “Building Your First AI Agent with Make.com” immediately feels more polished when paired with:
- A relevant cover image
- A clear, customized summary
- A presentation that looks more like a finished resource than a rough workspace
That is a simple change, but a useful one.
Chrome Skills: Turn Your Best Prompts Into One-Click Tools
This may actually be one of the most practical updates in the bunch.
Inside Google Chrome, you can now save prompts as reusable skills. In other words, if you have a prompt you use over and over again, you can convert it into a quick-access tool instead of rewriting it every time.
How Chrome Skills Work
The workflow is simple. You type a prompt and use backslash to begin turning it into a skill. Then you can:
- Save the prompt as a skill
- Give it a name
- Review or edit instructions
- Optimize the prompt
You can also access a skills library where you browse your saved skills and explore available ones as Google expands the system.
The Current Limitation
Right now, this feature is only available in Chrome. It is not yet built directly into Gemini in the same way.
So if you want reusable prompt management across Gemini, ChatGPT, and Claude, a dedicated prompt tool can still make sense. But Chrome skills are a really interesting sign of where Google is headed. The browser is becoming an AI workspace, not just a place to open tabs.
Why Skills Matter for Productivity
The best prompts are usually not one-offs. They are repeatable workflows. That is why this feature is so valuable.
You can turn common tasks into one-click tools, such as:
- Summarizing research notes
- Generating content briefs
- Rewriting emails
- Creating first-draft social posts
- Extracting action items from documents
Anything you do repeatedly with AI becomes a candidate for a skill.
The New Gemini Mac App Is the Foundation for a Super App
Google also released a Gemini app for Mac, and while it looks fairly basic today, this is one of the updates worth paying the closest attention to.
Right now, the app mostly mirrors what you can already do in Gemini on the web. But the structure of the app makes it obvious that Google is building toward something much bigger.
What Is Already Inside the Gemini Desktop App
In the current Mac app, you can access:
- Memory
- Connected apps
- Custom instructions
- Keyboard shortcuts and controls
Under the “more tools” area, the app includes access to:
- Personal intelligence
- Guided learning
- Deep research
- Canvas
- Music creation
- Video creation
- Image creation
- NotebookLM
- Google Photos
- Google Drive
- Files
That is already a pretty broad toolset. But the most important feature may be the ability to share a window with Gemini directly from the app.
Screen Sharing Is the Feature to Watch
The app lets you share your screen or a window with Gemini, which opens the door to a much more active assistant experience.
That is significant because it moves Gemini closer to the desktop assistant model that other major AI platforms are pursuing. The direction is clear: these tools are evolving from passive chatbots into systems that can observe context, assist in real time, and eventually take action.
Why This App Matters Even If It Feels Basic Today
At the moment, the Gemini Mac app does not do anything radically different from the web version. But that is not the point.
The point is that Google now has the shell for a true Gemini super app. As agents roll out, this desktop app is likely to become one of the most important parts of the Gemini ecosystem.
If you are trying to automate tasks, increase productivity, or work more fluidly across files, windows, and apps, this is the release to keep an eye on.
The Bigger Trend: AI Companies Are Building Operating Systems, Not Just Chatbots
One of the clearest takeaways from these updates is that the major AI platforms are all converging on a similar strategy.
They are no longer just improving models. They are building full environments around those models:
- Specialized creative tools like Flow Music
- Desktop apps that can eventually act on your behalf
- Reusable skills and workflows inside the browser
- Integrated productivity ecosystems tied to files, memory, and connected apps
That is why these Gemini updates matter. Even the smaller changes point toward a future where AI is not something you occasionally open. It becomes part of how you create, organize, research, and execute throughout the day.
Practical Takeaways
If you want the short version of what is worth trying first, here is where I would focus:
- Test Flow Music if you do anything with content, audio, or creative ideation.
- Upgrade AI Studio access if you are a serious Gemini power user and want better limits and model access.
- Use NotebookLM covers and summaries if you share notebooks professionally.
- Create Chrome skills for prompts you repeat all the time.
- Install the Gemini Mac app now so you are ready as more advanced desktop agent features arrive.
Suggested Media to Include With This Article
To improve engagement and SEO, this article would pair well with a few visuals:
- Screenshot of Flow Music session interface
Suggested alt text: “Google Flow Music app interface for AI song and beat generation” - Screenshot of NotebookLM cover customization
Suggested alt text: “NotebookLM custom cover image and notebook summary settings” - Screenshot of Chrome skills creation flow
Suggested alt text: “Google Chrome prompt saved as reusable AI skill” - Screenshot of the Gemini Mac desktop app tools menu
Suggested alt text: “Gemini Mac app with deep research canvas and screen sharing tools”
FAQ
What is Google Flow Music?
Flow Music is Google Gemini’s new music creation app. It allows you to generate beats, vocals, riffs, songs, mixes, and even music videos using prompts, uploaded media, and conversational editing.
Can Flow Music use images or recorded audio as input?
Yes. You can record live audio into a session, upload an image to inspire lyrics, or upload your own audio to guide the creative process.
What changed in Google AI Studio?
AI Studio now makes it easier to upgrade access using either a Gemini API key or a Google AI subscription. The API key route gives access to all models and agents, while the subscription provides higher limits and additional Google AI benefits.
What are the new NotebookLM features?
NotebookLM now supports custom notebook covers, titles, and summaries. These features are especially useful when sharing notebooks with other people because they improve presentation and context.
What are Chrome skills in Gemini?
Chrome skills let you save useful prompts as reusable tools directly inside Google Chrome. You can name them, optimize them, and access them from a skills library for faster repeated workflows.
Is the Gemini desktop app available for Mac?
Yes. Google released a Gemini app for Mac. It currently mirrors much of the web experience, but it already includes access to several tools and supports window sharing, which points toward more advanced desktop agent features in the future.
Final Thoughts
Some AI updates are minor interface tweaks dressed up like major launches. This is not that. Google Gemini’s new features show a real expansion in capability and direction.
Flow Music is the standout release right now because it feels fresh, ambitious, and immediately fun to use. Chrome skills may end up being the sleeper hit because they solve a real workflow problem. And the Gemini Mac app, even in its current form, looks like the beginning of something much bigger.
If you are building your AI toolkit for the next year, these are exactly the kinds of releases worth paying attention to.
If you found this useful, share it with someone exploring AI productivity tools, and check out related guides on Gemini, NotebookLM, and prompt workflows to stay ahead of what Google is building next.
