The Apple AirPods Max 2 arrive in a wireless headphone category that has become far more competitive since the original model launched. Sony, Bose, and Beats have all continued to refine ANC, comfort, battery life, app control, and sound quality, which makes small updates harder to justify at premium pricing.

For existing Apple users, the improvements may be enough: better noise cancelling, tighter ecosystem integration, and slightly improved sound quality. But for anyone still using an older pair of wireless headphones, or considering a move from Sony, Bose, or Beats, the AirPods Max 2 need to offer a clearer reason to switch. Marginally better is still better, but in 2026, it may not be enough. How do they compare to the model they replace?

Physically, these are almost identical headphones: same industrial design, same aluminum construction, same weight, same controls, and the same overall sound signature. Even the improvements, including better ANC, USB-C wired audio, and a slightly firmer low end, feel deliberately restrained rather than transformative.

Compared directly against the original AirPods Max, the overall experience remains remarkably similar. Apple clearly viewed this update as a refinement rather than a reinvention.

Design & Features 

The AirPods Max 2, like the original, still feel like one of the most premium wireless headphones Apple has ever made. The aluminum ear cups feel excellent, the suspension headband helps spread the weight better than expected, and Apple’s Transparency, Noise Cancelling, and Adaptive modes remain among the best implementations in the category.

That weight still matters. At 382.2 grams, the AirPods Max 2 never disappear on your head. The balance is good, but after a full record, taking them off feels less like a break and more like your neck filing a polite complaint.

The listening modes are very well executed. Noise Cancelling is excellent with steady background noise and also reduces intermittent sounds effectively. Transparency mode keeps outside sound natural and useful, while Adaptive mode works as a smart middle ground, adjusting continuously based on the listening environment.

Sound

The AirPods Max 2 retain almost all of the strengths (and occasional weaknesses) of the original model. The stereo image still sounds expansive, the low end still has satisfying weight and impact, and the headphone’s detail retrieval remains impressive for a wireless headphone.

The general tuning philosophy seems unchanged as well: a mostly inoffensive sound with mild elevation in the upper treble and bass, nicely filled-in lower mids that give vocals a nice weight and intimacy, and a laid-back lower treble that remains forgiving across a wide range of music and other content.

Reference Tracks

Bass

The low end comes across as classic Apple. There is a typical elevation to the bass and sub-bass that gives kick drums and bass guitars an enhanced sense of impact and authority.

Listening to “Only When I” by Alice Phoebe Lou, despite the relatively emphasized low end, the separation between kick drum and bass guitar remains nice and distinct, letting each support the arrangement within its own range of low-end frequencies.

Listening to “Escalator” by Ritt Romney, the AirPods Max 2 do a pretty great job reproducing both the sustained sub-bass and pulsating mid-bass rhythm without the two becoming overly blurred together.

While the low end is not as punchy or authoritative as it is on something like a planar magnetic open-back driver Audeze LCD-X or a dynamic open-back like a Focal Clear, it sounds just as good or better than any other closed-back wireless headphone I’ve heard without sacrificing midrange intelligibility. 

Midrange

While the low end on the AirPods Max 2 lands decidedly north of neutral, the midrange plays it very safe.

The lower midrange where the fundamental frequencies of melodic elements like vocals, guitars, and pianos sit, is more forward and filled out. Moving into the upper midrange from roughly 1 kHz to 5 kHz, Apple has done the typical Apple thing and kept the amplitude nice and relaxed.

This is where overemphasis becomes most irritating because human hearing is especially sensitive in this range. It is where many urgent sounds live, including babies crying, people screaming, and alarms.

The downside of playing it safe is that some songs and content come across as overly dull or relaxed. One example is “Halloween” by Phoebe Bridgers. The inherently forward lower midrange combined with the relaxed upper midrange of the track makes for an especially wonky rendition of a mix that is usually one of my favorites. The buildup in the upper bass and lower midrange creates a sense of pressure and fullness that needs more upper midrange clarity to keep things balanced.

Generally, Apple did a good job sculpting the midrange to be widely palatable across most music and audio content.

Treble

The treble, like the low end, is somewhat emphasized. Fortunately, Apple mostly avoids the kind of sharp lower-treble peaks that can make headphones sound brittle or fatiguing over longer listening sessions.

Instead, most of the added energy seems concentrated higher up in the spectrum, lending cymbals, acoustic textures, reverbs, and ambient details a pleasant sense of air and openness without constantly shoving them into the foreground.

The treble emphasis does occasionally land unpleasantly with sibilance and percussion, depending on how a song is mixed.

Listening to “Fruity” by Rubblebucket, the vocal sibilance rides dangerously close to the edge of unpleasantness without ever fully crossing over. Songs with more aggressive mixes like “Life or Just Living” by Caveman, fall squarely on the side of unpleasant every time the vocalist sings anything sibilant or consonant. 

Overall, Apple did a decent job making the AirPods Max 2 sound exciting without becoming overly fatiguing. Some customization would go a long way, but as usual, Apple provides no native EQ control.

Comparison to AirPods Max 1

While the tuning feels more similar than different compared to the original AirPods Max, the upper and lower extremities of the frequency range come across as subtly firmer and more defined. Bass sustain feels more convincing, and upper-treble information like percussion and cymbals sounds smoother and slightly less digital.

Plugging the AirPods Max 2 in via USB-C for wired listening also improves the sound modestly, further refining many of the qualities above. The difference is not night and day, but for someone seeking the most natural sound possible from the AirPods Max 2, listening via USB-C is the best way to get there.

The improvements to noise canceling and transparency functions are similarly incremental. If you compare the two models side by side, the newer model is clearly superior in overall attenuation of environmental noise. The AirPods Max 2 do a more even job across the frequency spectrum, whereas the original model lets slightly more lower-treble information through to the listener.

However, the difference is not night and day. The original already did a fantastic job and the new version only slightly improves on the former.

For anyone deciding between the old version and the new version, I would suggest the AirPods Max 2 as the superior product. But is the upgrade from the original worth the money? I don’t think so. The two products are more similar than different so I would wait for the next iteration.

Comparison to Bose QuietComfort Ultra 2

The Bose QuietComfort Ultra (2nd Gen) feels like the more rational product. It is lighter, more comfortable, has notably better battery life, folds up more easily and compactly, and includes native EQ sound and noise cancelling level customization. Ultimately, the Bose is a more practical design for listening on the go.

Next to the Bose, the AirPods Max 2 feels more like an Apple luxury fashion accessory. The materials are nicer, the integration is more seamless with my iPhone, and all three listening modes (Transparency, Noise Cancelling, Adaptive) sound more natural despite the lack of customization. 

In terms of sound, the Bose QuietComfort Ultra 2 sounds surprisingly different compared to the AirPods Max 2 for two headphones competing in the same space. The Bose takes Apple’s V-shaped frequency response and puts it into warp drive – both the low bass and the upper trouble are more emphasized on the Bose. Ultimately, I find the QC Ultra 2 less detailed and more fatiguing than the AirPods Max 2 with a generally more artificial sound.

However, between the QC Ultra II’s significantly lower price, lighter weight, and inclusion of EQ customization, I would sooner recommend the Bose to someone looking for a solid sounding over-ear noise canceling headphone.

The Bottom Line

The AirPods Max 2 are, ultimately, a product for people who already know they want an AirPods Max.

If you’re deeply invested in Apple’s ecosystem, value best-in-class transparency mode and noise cancellation, care about industrial design, and want a wireless headphone that sounds genuinely good without requiring much thought or tweaking, the AirPods Max 2 stands out as one of the most compelling options on the market.

They still feel uniquely Apple in both the best and worst ways: beautifully built, thoughtfully integrated, and exceedingly expensive.

At the same time, this update feels particularly conservative. The AirPods Max 2 improves on the original model in nearly every category – but only slightly. The sound is more refined, the ANC is more effective, and USB-C wired audio is a genuinely welcome addition. But none of it fundamentally changes the experience. Existing AirPods Max owners are not missing much.

Personally, there are very few situations where I would prefer listening to the AirPods Max 2 over my much-loved AirPods Pro 2 earbuds. Between the AirPods Max 2’s less-neutral sound signature and dramatically heavier weight, I simply get along better with the sound and form factor of the AirPods Pro 2.

If my goal is to listen to headphones with the best possible sound quality, I prefer wired open-back headphones. If convenience is the mission, I’m always going to choose the pocketable AirPods Pro 2. Maybe if Apple were to add some EQ customization to the AirPods Max 2, this conclusion may have looked very different but based on Apple’s track-record, I won’t hold my breath.

Pros

• Best-in-class Transparency, Noise Canceling, and Adaptive modes
• Spacious and detailed sound
• USB-C audio support
• Apple ecosystem integration

Cons:

• Below average battery life
• Heavy
• No built-in EQ customization
• Expensive!

Ratings Compared

Where to buy:

Related Reading:

Looking for the most recent Mini Crossword answer? Click here for today’s Mini Crossword hints, as well as our daily answers and hints for The New York Times Wordle, Strands, Connections and Connections: Sports Edition puzzles.

Need some help with today’s Mini Crossword? The clue that threw me off was 7-Across, but I eventually figured it out. Read on for all the answers. And if you could use some hints and guidance for daily solving, check out our Mini Crossword tips.

If you’re looking for today’s Wordle, Connections, Connections: Sports Edition and Strands answers, you can visit CNET’s NYT puzzle hints page.

Read more: Tips and Tricks for Solving The New York Times Mini Crossword

Let’s get to those Mini Crossword clues and answers.

Mini across clues and answers

1A clue: Not such a wild place to see an elephant
Answer: ZOO

4A clue: Mark who wrote “Man was made at the end of the week’s work, when God was tired”
Answer: TWAIN

6A clue: Symbol on the U.S. quarter
Answer: EAGLE

7A clue: Parts of a dental exam
Answer: XRAYS

8A clue: What Wile E. Coyote is always exploding himself with
Answer: TNT

Mini down clues and answers

1D clue: Big name in restaurant ratings
Answer: ZAGAT

2D clue: Like your hands after eating potato chips, perhaps
Answer: OILY

3D clue: Lowest numbers in sudoku
Answer: ONES

4D clue: Reach out with your fingers?
Answer: TEXT

5D clue: Give an alert
Answer: WARN

At Google I/O, the company unveiled Managed Agents in its Gemini API — a service that promises to collapse weeks of agent deployment work into a single API call. It’s also a sign that Google believes its ecosystem, including the newly launched Antigravity CLI, is ready to own the execution layer end-to-end.

Before a single agent is written, teams are already spending days on the unglamorous work: standing up execution environments, managing sandboxes, wiring tool call infrastructure. Model providers like Anthropic have launched platforms to handle much of that work — but Google’s approach is different.

Google said in a blog post that Managed Agents in the Gemini API abstracts “away the complexity so that you can focus on your product experience and agent behavior.” The service is available in preview via new custom templates in Google AI Studio.

The growth has introduced a real architectural question: should agent management live at the execution layer — embedded in the model or its harness — or at the infrastructure layer, as a separate runtime?

Comparing Google’s approach

Until recently, agent orchestration relied on frameworks that sat above the model, directing agents and letting teams control routing and execution separately. That layer is now being absorbed by the platforms themselves.

Recent platforms like Claude Managed Agents embed orchestration at the model layer rather than on a separate runtime platform. The idea is that the model owns the reasoning and orchestration layers, and enterprises have control over execution. 

AWS, through new capabilities on Bedrock AgentCore, adds managed harnesses that stitch together the upfront tasks for deploying agents. Google’s approach goes further, optimizing the model, harness, and sandbox together and running everything in secure Google-managed environments.

René Sultan of Ramp, cited in Google’s announcement, said the shift is concrete: “The real shift with Gemini Managed Agents is that the agent runtime moves into the platform. With the sandbox, infrastructure and execution loop managed for you, developers can focus on productizing the agent’s domain-specific behavior and iterating at a completely different pace.”

The new orchestration reality 

Enterprises starting fresh with agents could find the platform offerings from Anthropic and Google strong, especially since they remove much of the difficulty of deploying agents while still maintaining some control. Google, however, is pushing for a more vertically integrated system, while Anthropic is betting on the model layer as an orchestration plane, and AWS focuses on authorization. 

But this also brings some risks, according to XYO founder and chief executive Arie Trouw.

“An additional risk is that developers will switch out what previously were deterministic services for what will now be probabilistic services, which can introduce unpredictable outcomes for the users at best, or data corruption at worst,” Trouw told VentureBeat in an email. “This is the classic example of having an amazing hammer and everything starting to look like nails. I’ve seen this pattern repeatedly as a developer and business founder myself in the past few decades.”

Scientific papers rely on readers trusting their information. That’s why it’s disturbing that a new study by researchers connected with Cornell and UCLA found 146,900 AI-generated fake citations in scientific papers hosted across four major research databases. 

A key limitation of large language models such as Gemini and ChatGPT is their tendency to produce plausible-sounding but incorrect information, a phenomenon known as hallucination. If a researcher relies on a chatbot to draft citations without verifying them, the model may generate references that are entirely fabricated.

While scientific papers are often hidden from the public eye, the research they report has a profound impact on our lives. Everything from the internet to lithium-ion batteries began as a research paper. 

But when scientists submit papers that cite AI hallucinations, it can erode faith in the quality of the research. 

Sloppy science

The research team analyzed 111 million references from 2.5 million scientific papers. They looked for citations with titles that the team could not match to any publication. While some of these instances were just spelling errors, the team also found hallucinations. 

Unscrupulous researchers had faked citations long before the rise of chatbots, so the team also examined the rates of unmatched citations in research published before 2023, when chatbots hadn’t yet become ubiquitous. 

“We find a sharp rise in non-existent references following widespread LLM adoption,” the authors write in the paper. 

The team also found that the bad citations were spread across many papers rather than concentrated in just a few. That suggests the problem is widespread, with many researchers relying on AI-generated references without fully verifying them.

Warning signs

Usha Haley, professor of management at Wichita State University, told CNET via email that she sees the proliferation of fake citations as a serious warning.

“Fake or AI-generated citations undermine trust in the scholarly record that provides the foundation on which peer review and cumulative knowledge rest,” Haley said. “Disturbingly, this skepticism is now coming from within academia itself and from early career scholars.”

The four databases where the researchers found the fake citations are arXiv, bioRxiv, SSRN and PubMed Central. These organizations, known as scientific repositories, play a major role in the research world. 

Before a paper is published in a scientific journal, the authors often upload it to a scientific repository, increasing its visibility and allowing the global scientific community to access it immediately. The new paper on AI hallucinating citations is currently hosted on arXiv.

Recently, arXiv has taken steps to stem the flow of false citations. The organization announced Tuesday that it will ban authors who submit work with hallucinated citations or with any sign of AI content that hasn’t been carefully checked. 

“The corpus of science is getting diluted. A lot of the AI stuff is either actively wrong or it’s meaningless. It’s just noise,” arXiv scientific director Steinn Sigurdsson told CNET’s Katelyn Chedraoui back in February. “It makes it harder to find what’s really happening, and it can misdirect people.”

Nvidia founder and CEO Jensen Huang is, perhaps, one of the greatest corporate hype men of all time when it comes to his company. He may even surpass Salesforce’s Marc Benioff when it comes to relentless optimism in his company’s future and revenues.

Even so, he delivers on the hype, quarter after quarter.

Instead of cautioning you to view the proclamation that he’s found a “brand new $200 billion TAM for Nvidia” with skepticism, I’d argue he’s earned a bit of trust.

Huang positioned this massive new market at the feet of Nvidia’s new CPU product, Vera, which was introduced in March. Speaking on Wednesday’s earnings call — after Nvidia posted another record-breaking quarter with $81.6 billion in revenue and forecast $91 billion for the next — Huang pitched Vera as a potentially transformative product. And one that already has promising sales figures.

But no matter how well Nvidia delivers, Wall Street harbors anxiety over what will knock Nvidia from its perch.

Lately, such fears have centered on the CPU. Nvidia is the king of the GPU, whereas historically the CPU markets were owned by companies like Intel and AMD. (Nvidia has made CPUs previously, of course, but that’s not its core business.)

For example, last month Amazon Web Services crowed about a giant contract it signed with Meta for millions of Amazon’s homegrown AI CPUs. Amazon CEO Andy Jassy has been clear that he thinks AWS can do AI chips, both GPUs and CPUs, at least as well, and possibly better than Nvidia.

But now, with the Vera CPU, which is sold alone and bundled with its Rubin GPU, Huang believes he’s unlocked “a major new growth driver” for his company because Vera is, he believes, “the world’s first CPU, purpose-built for agentic AI,” Huang said on the call.

“Vera opens a brand new $200 billion TAM for Nvidia, a market we have never addressed before, and every major hyperscaler and system maker is partnering with us to deploy it. The world is rebuilding computing for agentic AI and robotic physical AI. Nvidia sits at the center of these transitions,” hype man Huang said.

He explained that while the “thinking” part of an AI model uses GPUs, agents mostly run on CPUs. They use CPUs to do their assigned tasks and will, he predicts, run their own form of CPU-driven PCs.

Vera is for agents because it’s specifically designed to process tokens as fast as possible. This is opposed to classic cloud architecture CPUs designed with “cores,” or the ability to run multiple instances of apps as fast as possible.

That sounds logical, but with the major cloud providers as well as startups pursuing AI chip development, what makes him think that Nvidia will be the go-to source for agentic CPUs?

Because, Huang says, Nvidia has already sold $20 billion worth of standalone Vera CPUs this year and we’re only at the beginning.

“The world has a billion users, human users. My sense is that the world is going to have billions of agents, not today. I mean, we’re going to grow into it, but we’ll have billions of agents, and those billions of agents will all use tools. And those tools are going to be like PCs, just like us humans using using PCs today,” he said.

“We’re going to need a lot more CPUs,” he explained.