NotebookLM Becomes Gemini Notebook, Claude Code Defaults to Sonnet 5, Suno Gets Cover Art Control: August 2026 AI Digest

In the first week of August, three AI tools quietly shipped meaningful updates.

Google officially rebranded NotebookLM as Gemini Notebook, integrating it fully into the Gemini ecosystem while rolling out code execution to all Pro users. Anthropic switched Claude Code's default model to Claude Sonnet 5 and moved code review out of the main conversation thread into a background subagent. Suno added a Duration slider that lets you set song length before generating, and a text-prompt-driven cover art editor that lets you iterate toward the visual you want.

As an EdTech CEO and AI tools researcher, here is what each update actually changes in practice.

Gemini Notebook: NotebookLM Gets a New Name and Runs Code
Claude Code + Sonnet 5: Default Model Upgrade and Review Offloading
Suno V5.5: AI Takes Control of Length and Visuals
What These Three Updates Tell Us

1. Gemini Notebook: NotebookLM Gets a New Name and Runs Code

On July 16, 2026, Google renamed NotebookLM to Gemini Notebook.

This is more than a cosmetic change. Google has fully absorbed NotebookLM into the Gemini brand and paired the rebrand with substantive capability expansion. The app now lives at notebook.google.com — the old address redirects automatically.

![Gemini Notebook — new blue-purple gradient logo and code execution interface]

What Actually Changed: Code Execution for All Pro Users

NotebookLM was always strong at analyzing and summarizing documents. What it could not do was execute anything. That changes now. Every notebook comes with an isolated "secure cloud computer." Type "analyze the trends in this spreadsheet" and Gemini Notebook writes Python, runs it server-side, and returns the results. No coding required.

A teacher can upload a class grade spreadsheet and ask "show me the grade distribution this semester" — and receive a chart and analysis without writing a single line of code.

Feature	NotebookLM (Before)	Gemini Notebook (After)
App URL	notebooklm.google.com	notebook.google.com
Code Execution	Not available	All Pro users
Logo	Previous design	Blue-purple Gemini gradient
Audio Summary	Standard Audio Overview	Customizable length, focus, and complexity
Interactive Mode	Passive listening	"Raise hand" for real-time questions

Audio Overview Gets Smarter

Audio summaries have been upgraded too. Previously, NotebookLM chose the format and length automatically. Now you can specify how long the summary should be, what it should focus on, and how technical it should be. The new Interactive Mode lets you "raise your hand" while listening to ask a question in real time — effectively interrupting the AI podcast hosts to dig deeper.

Tip: Try specifying both length and difficulty when generating Audio Overviews. "Summarize this research paper as a 10-minute audio briefing — avoid jargon and use plain language" yields far more useful results than a generic prompt. Code execution is currently live for Google AI Ultra subscribers and Workspace business customers; Pro web rollout is ongoing.

2. Claude Code + Sonnet 5: Default Model Upgrade and Review Offloading

In August 2026, Claude Code's default model switched to Claude Sonnet 5.

Anthropic gave Claude Sonnet 5 a native 1-million-token context window and made it the built-in default for Claude Code. Through August 31, promotional pricing applies: $2 per million input tokens and$ 10 per million output tokens. In practical terms, you get substantially more capability for the same cost.

![Claude Code + Sonnet 5 — new default model interface showing 1M-token context window]

Code Review Moves to the Background

The update that changes the daily experience most is how /code-review now works. It runs as a background subagent instead of inside the main conversation thread.

Previously, triggering /code-review blocked the conversation — the entire review process played out in your main window, making it hard to continue other work, and filling your history with review output. Now the review runs separately. You can start a review and immediately keep writing code, then check the results when it finishes.

Enterprise TLS Proxy: Zero Extra Setup

For teams running Claude Code behind corporate network proxies, this update removes a common friction point. Claude Code now trusts the operating system's CA certificate store by default. Enterprise TLS proxies that previously required manual certificate configuration now work out of the box.

/recap: Re-orient Instantly When You Return

Long sessions can be disorienting after a break. The new /recap command generates a contextual summary of the current session on demand. You can also configure it in /config to trigger automatically whenever you re-enter a session. Instead of scrolling back through history to remember where you left off, a single command brings you up to speed.

Tip: If you work on large codebases, make /recap your session-start habit. Pair it with the 1M-token context window to keep far more project context in play. For reviews, start /code-review on a long file, immediately switch to another task, and return to the review output when it's ready — your main conversation stays clean throughout.

3. Suno V5.5: AI Takes Control of Length and Visuals

In late July and early August, Suno added two features that complete different gaps in the creation workflow: a Duration slider and a text-based cover art editor.

Just two weeks after rebuilding the Lyrics Editor from scratch, Suno has addressed two more fundamental constraints: how long a track is, and what it looks like.

![Suno V5.5 — Duration slider and cover art text-prompt interface side by side]

Duration Slider: Set the Length Before You Generate

Released July 20. The Create form on the web now includes a Duration slider for V5.5-model generations. You pick the exact length you want before hitting generate.

Before this, Suno chose the song length autonomously. If you needed a 30-second background loop for a video or a 15-second podcast intro, you regenerated repeatedly and hoped. Now you set it once and get what you asked for.

Use Case	Recommended Duration
Video intro	10–15 seconds
Classroom background	2–3 minutes
Podcast outro	30–45 seconds
Full streaming track	3:30–4:00 minutes

Cover Art Editor: Iterate With Text Prompts

Released July 31. Suno's web interface now lets you generate cover art via text prompt and refine it iteratively. The feature works for songs, playlists, and profile images, and supports both static images and short video loops.

Previously, Suno auto-generated cover art from your lyrics and genre. If it missed the mark, your only option was "regenerate and hope." Now you can write "darker and more dramatic mood" or "add a night sky with stars in the background" and steer the output in the direction you want.

Tip: Combine style reference and composition in a single cover art prompt — "movie poster style, solid dark background, bold white serif font, artist name at top" — to reach your target in one or two iterations rather than ten. Remember to select the V5.5 model to unlock the Duration slider. Set both length and cover art prompt before generating for a fully finished track in one pass.

4. What These Three Updates Tell Us

There is a shared direction in all three updates.

The tool is no longer just executing what you tell it — it is taking over entire sub-tasks. Gemini Notebook analyzes your data instead of telling you how to analyze it. Claude Code's /code-review runs the review and surfaces only the results. Suno refines your visual direction instead of generating blind.

From an EdTech perspective, this shift carries an important implication for both students and teachers. As AI tools take on more execution, the critical skill is no longer only "how to use the tool" — it is "how to evaluate what the AI produced." The narrower the human input becomes, the sharper the human judgment needs to be.