Meta description: Learn the latest NotebookLM and Google Gemini upgrades: insert notebooks into Gemini chats, create cinematic videos, import memory, and build Workspace automations.
Google just pushed a stack of upgrades across NotebookLM and Gemini, and the big theme is simple: your AI workflows are getting faster, more flexible, and more connected to the rest of the Google ecosystem. If you rely on these tools for writing, research, content creation, or internal productivity, the changes can save you a surprising amount of time.
This guide breaks down what’s new and, more importantly, how to use it. You will learn how to: bring your NotebookLM notebooks directly into Gemini chats, get higher control over what the AI generates, create cinematic “video overview” content, configure smarter responses, import memory and chats into Gemini for personalization, and build automations with Workspace Studio. We’ll also touch the newest model options in Google AI Studio, including “flash live” for low-latency conversation and a cheaper VO3.1 light video model.
Table of Contents
- 1) Use NotebookLM notebooks inside Google Gemini chats
- 2) Add notebooks to “Gems” in Gemini
- 3) NotebookLM video overviews now support cinematic videos
- 4) Configure how NotebookLM chats respond (roles, length, and behaviour)
- 5) A better infographic experience
- 6) Import memory from Gemini and import other chats into Gemini
- 7) Build AI automations with Workspace Studio
- 8) New models in Google AI Studio (live, offline-friendly, and cheaper video)
- Practical workflow ideas you can try today
- FAQ
- Recommended next step
1) Use NotebookLM notebooks inside Google Gemini chats
One of the most practical upgrades is that you can now insert NotebookLM notebooks directly into Gemini. Instead of switching between tools to get the job done, you can place your notebook context into any Gemini chat and ask for outputs based on that notebook.
What changed in Gemini
When you are in Gemini and click the plus (add attachment), you now have upload options. Under the “more uploads” area, you will see Notebooks. That means you can select a notebook you created in NotebookLM and insert it into a chat.
Why this matters
The key advantage is speed and control. The same notebook content can be used as a foundation, but Gemini can produce structured outputs more quickly and with more direct instruction.
Here’s the type of workflow that becomes much easier:
- Start with a NotebookLM notebook (for example, a research pack or a guide).
- Insert that notebook into a Gemini chat.
- Ask Gemini to produce specific deliverables, like a research paper plus an image for a cover plus an SEO-optimized blog post.
In practice, this is the difference between “the notebook can generate content” versus “I can command the notebook’s knowledge to generate multiple assets in one go, inside a chat I control.”
NotebookLM vs Gemini for different tasks
NotebookLM is great for grounded work, but some output types and formatting controls can feel limiting. Gemini, on the other hand, often completes those tasks more quickly and can follow more varied instructions in the same session.
With NotebookLM alone, you might be able to generate certain media like infographics, or request reports, but it may take longer or provide less direct control over other formats. With the new “insert notebook into Gemini chat” approach, you can request what you need more directly.
2) Add notebooks to “Gems” in Gemini
It is not just standard chat. You can also add NotebookLM notebooks into Gems on Google Gemini.
The workflow described is:
- Create or open a Gem.
- Go to Knowledge.
- Click the plus under knowledge to add your NotebookLM notebooks.
This effectively lets you reuse your NotebookLM knowledge inside a more persistent Gemini setup. Instead of recreating context each time, you can build a Gemini “assistant” that is grounded in the notebook content you care about.
Suggested next step: If you have notebooks you reuse often (industry research, product documentation, weekly planning, prompt libraries), convert them into Gem knowledge so you can consistently generate outputs faster.
3) NotebookLM video overviews now support cinematic videos
NotebookLM also gained a major media upgrade: “video overview” now creates cinematic videos, not just slide-style sequences.
What you get now
Inside a NotebookLM notebook, select the video overview option. You will see an updated experience labeled as cinematic videos. The output described is much more than a basic slideshow with audio.
Instead, it uses generated motion graphics built from the notebook’s images, combined with an overview style presentation.
How it’s different from before
Earlier video creation tended to function like a slideshow: images advancing step by step with an audio summary. Now it focuses on cinematic effects and motion graphics, making the final output feel more like a polished piece of content.
In the example provided, the system took images and made them motion graphics using a VO3.1-based approach (the transcript mentions VO3.1 explicitly).
4) Configure how NotebookLM chats respond (roles, length, and behaviour)
NotebookLM now gives you more control over how the chat behaves. This is especially useful when the same notebook content needs different tones or output styles for school versus work.
Custom instructions for different projects
You can configure the chat to respond in a specific way depending on the context. For example:
- School project: more explanation, clearer structure, study-friendly formatting.
- Work project: more concise results, professional tone, decision-ready formatting.
Add “roles” to steer outputs
The update also includes the ability to set roles. You can ask NotebookLM to “pretend to be” a specific persona or specialist, which nudges its style, vocabulary, and assumptions.
Adjust response length
You can also choose whether responses should be longer or shorter. This is a surprisingly big quality-of-life improvement because it reduces the back-and-forth of “make it shorter” or “expand this.”
5) A better infographic experience
Infographics are often where “AI output” stops being useful if you cannot guide the style and detail. NotebookLM’s upgrade improves that.
When you generate an infographic now, you can choose:
- Visual style (the format and layout direction)
- Level of detail (how deep it goes)
- Orientation (how it is presented visually)
That gives you more control over whether the infographic is a quick summary asset or a more information-dense graphic for reports and content marketing.
6) Import memory from Gemini and import other chats into Gemini
This is one of the most “power-user” upgrades: Gemini can import memory, and Gemini settings can be used to bring in context from other AI tools.
Import memory from Gemini (for use in other tools)
In settings, there is now an option for import memory from Gemini. The system gives you a prompt you can copy and paste into another AI. For example, you could paste it into Claude or another chat system, and it will respond using Gemini memory context.
This is useful if you have a workflow that depends on a different model for certain tasks, but you still want the personalization or memory to carry over.
Import chats exported from other platforms
Another big feature is importing chat history. The described flow is:
- Export your data from another AI (example: ChatGPT data export).
- In Gemini, add the exported zip file.
- Gemini uses it to build more context about your interactions.
The practical benefit is clear: Gemini can become more aware of how you communicate, what you ask for, and the kinds of preferences you show through past conversations.
Make sure personalization toggles are enabled
To get the full effect, you want to ensure settings like personalized intelligence and connected apps are enabled.
The described checklist includes:
- Memory features turned on under personalized intelligence
- Connected apps turned on
- Relevant instruction settings for Gemini enabled
Search and YouTube context can matter
One key setting mentioned is access to how you interact with Google Search, YouTube, Google News, Google Shopping, Google Maps. Turning these on gives Gemini richer context for personalized responses.
7) Build AI automations with Workspace Studio
If you want more than “ask AI and get text,” the update to Workspace Studio is where things get operational.
The platform supports automation building with Google Gemini-powered actions. There are two ways to get started:
- Discover: describe a test or desired workflow and have the system generate the agent, connectors, and flow structure
- Templates plus custom building: start from ready-made automation patterns, then refine or build your own
Examples of automations described
The transcript describes automations like:
- Automatically catching you up on what happened “yesterday” on a schedule
- Email follow-up tasks
- Daily summary of unread emails
- Notifications about emails from key people
Then it goes further: you can trigger flows when forms are filled out, and you can add actions in Gemini that summarize, extract, and recap information.
Connect actions across Google tools
Because Workspace Studio sits in the Google ecosystem, automations can interact with:
- Gmail
- Chat
- Sheets
- Drive
- Docs
- Tasks
The described flow also supports logic like an end if statement, filtering, and branching to control what happens next.
Suggested reading: Explore Google’s Workspace automation tooling here: https://flows.workspace.google.com/
8) New models in Google AI Studio (live, offline-friendly, and cheaper video)
Beyond the user-facing product updates, Google also added model options in AI Studio that affect real-world performance and cost.
Gamma models for small devices and offline use
The transcript notes that there are new “gamma” models that can run on small devices and can also run offline. The main point is enabling AI interaction even when your phone or computer is offline (for example, airplane mode scenarios).
Gemini 3.1 Flash live preview
Google also released a Gemini 3.1 Flash live preview option. The advantages listed include:
- Better latency
- Better audio-to-audio optimization
- Real-time dialogue improvements like acoustics nuance detection
- Multimodal awareness (you can show it something and get a strong response)
VO3.1 light video model (lowest cost described)
For video generation, there is also VO3.1 light. The transcript calls out its cost comparison, stating it is the cheapest video model listed for Google at the time described.
Cost details mentioned:
- VO3.1 light: five cents
- VO3.1 fast: 15 cents
- VO3.1 3.1: 40 cents
Suggested reading: Learn more about Gemini models and API options in Google AI Studio: https://ai.google.dev/
Practical workflow ideas you can try today
With these changes, you can combine capabilities into workflows that feel way closer to a production system than “single chat prompts.” Here are a few examples built from the upgrades described:
From notebook research to SEO content faster
- Create your research notebook in NotebookLM.
- Insert that notebook into a Gemini chat.
- Ask Gemini to produce:
- a research paper style draft
- an image for the blog cover
- an SEO-optimized blog post version
- Refine formatting in Gemini without returning to NotebookLM repeatedly.
Turn a notebook into shareable multimedia
- Generate a NotebookLM video overview for your notebook.
- Use the new cinematic output as a social teaser or internal training piece.
- Pair with improved infographic generation for one post across multiple formats.
Personalize Gemini across tools by importing memory and chats
- Enable relevant memory and connected apps settings in Gemini.
- Import exported chat history from tools you already use.
- Use import memory to bring Gemini personalization into another model when needed.
Automate “busy work” with Workspace Studio
- Start with a template or use Discover to describe your goal.
- Choose scheduling, triggers (like forms submitted), and actions (summaries, extraction, notifications).
- Connect outcomes into Gmail, Docs, Sheets, or Tasks.
FAQ
Can I insert NotebookLM notebooks directly into Gemini chats?
Yes. In Gemini, click the plus button, then use “more uploads” to find notebooks and insert them into any chat so Gemini can generate outputs grounded in your notebook content.
What is new about NotebookLM video overviews?
NotebookLM now supports cinematic video overviews that go beyond a simple slideshow. It creates motion-graphics-style video content based on the notebook images, with an improved presentation experience.
Does NotebookLM let me customize the chat style?
Yes. You can configure response behaviour, including roles (persona-style prompts) and response length, and tailor instructions for different project types like school or work.
How does “import memory from Gemini” work?
Gemini settings provide a prompt you can copy into another AI tool. That prompt helps the other tool use Gemini memory context to generate responses aligned with your saved context.
Can I import chats from other AI platforms into Gemini?
Yes. You can export your data (for example, from ChatGPT), then add the exported zip file in Gemini so Gemini has more context from your prior conversations.
What can Workspace Studio automate?
You can build Gemini-powered automations across Google Workspace tools. Examples include scheduled email summaries, follow-up workflows, notifications from key people, form submission triggers, and actions in Gmail, Chat, Sheets, Drive, Docs, and Tasks.
Recommended next step
If you try one thing first, do this: pick a NotebookLM notebook you already have, insert it into Gemini using the new “Notebooks” upload option, and ask for multiple deliverables in one go (for example, a blog post plus a cover image). Then build from there.
If you have ideas for workflows you want to automate, leave a comment with your use case. Also consider sharing this with someone who uses Gemini or NotebookLM for writing, research, or everyday productivity. For more practical AI updates and workflows, explore related posts on your site that cover prompt engineering, Gemini tips, and Google Workspace automation.
