Qualcomm Unveils Two Snapdragon Chips Built to Push AI Deeper Into Phones

Qualcomm has introduced two new flagship smartphone processors, the Snapdragon 8 Elite Gen 6 and the Snapdragon 8 Elite Extreme Gen 6, as it doubles down on the idea that the next major battleground in mobile will be AI. Announced at the company’s annual Snapdragon Summit on September 22, the chips are designed to make on-device AI assistants smarter, faster, and more personal while also improving camera and audio performance.

The launch matters because Qualcomm is not just adding another round of incremental speed gains. The company is positioning its latest silicon as a foundation for a new generation of AI agents that can run locally on phones, store contextual memory, and handle voice tasks without always relying on the cloud. That could affect how consumers interact with their devices and how smartphone makers market their premium models over the next year.

What Qualcomm announced

Qualcomm’s new lineup consists of two premium mobile chips aimed at high-end Android phones. The standard Snapdragon 8 Elite Gen 6 is intended to deliver better AI efficiency and improved device performance, while the more powerful Snapdragon 8 Elite Extreme Gen 6 goes further with support for larger models and more advanced media features.

The company framed both processors as part of a broader shift toward smartphones that do more of the AI work directly on the device rather than in distant data centers. That approach can improve latency, reduce dependence on a constant network connection, and make certain personal features more private.

How do the new chips improve AI on phones?

They do it by expanding the hardware available for on-device inference and by adding a redesigned sensing hub that can run smaller AI models locally. Qualcomm said that the new sensing hub can handle models with up to 200 million parameters, enough to support features such as a personal scribe, speaker differentiation, and memory-based task suggestions.

In practical terms, the company is describing phones that can listen, organize, and anticipate user needs more effectively. A device could, for example, identify different speakers in a conversation, summarize what was said, and use prior behavior to make more useful recommendations or automate routine tasks.

Qualcomm said the updated sensing hub can support a full voice-in, voice-out agent directly on the phone, a sign that it expects local assistants to become a central selling point for premium mobile devices.

That same architecture is also meant to support what the company describes as more personalized AI agents. Rather than acting like generic chatbots, these assistants could build a memory of user preferences and context over time, helping the phone behave more like a persistent digital helper.

Why the Extreme version stands out

The Snapdragon 8 Elite Extreme Gen 6 is the most ambitious of the two chips, with the ability to run a 30-billion-parameter mixture-of-experts model locally. That is a significant step up in ambition for smartphone-class hardware and suggests Qualcomm is betting that future premium devices will be expected to run much larger models than today’s mainstream phones.

A mixture-of-experts, or MoE, model is designed so that only the parts needed for a specific task are activated at a given moment. This allows a model to be large overall without requiring all of its parameters to be used at once, making it more practical for mobile hardware than a dense model of the same total size.

That distinction matters because phone makers and chip designers are racing to find ways to run richer AI experiences without draining battery life or requiring constant cloud access. Qualcomm’s pitch is that the Extreme chip can push those boundaries further while still fitting within the constraints of a mobile device.

How does Qualcomm compare with Apple?

Qualcomm is aiming to match and, in some areas, surpass the kind of on-device AI capability that rivals have been showcasing. The company pointed to Apple’s Worldwide Developers Conference in June, where Apple announced a 20-billion-parameter mixture-of-experts model as part of its third generation of foundation models.

In other words, Qualcomm is trying to establish that its top-tier Android hardware can support local AI models at a scale beyond what Apple has publicly highlighted in its own mobile AI roadmap. That comparison is important because Apple and Qualcomm are both vying to define what high-end consumer AI should look like inside a phone.

The race is not simply about raw parameter count. It is also about how efficiently those models can be run, how much they improve the user experience, and how effectively they integrate with operating systems and apps. Still, the chip maker clearly wants buyers and handset partners to see the Extreme version as a serious AI platform, not just a faster processor.

What else do the chips improve besides AI?

Qualcomm said the new processors are also intended to improve camera quality, video capabilities, and audio processing. The company described the CPU as offering pixel-level camera control, better stabilization, and stronger motion understanding, all of which are meant to support more professional-looking photography and video capture.

On the video side, the Extreme version can support 8K recording at 60 frames per second and 4K recording at 240 frames per second, which opens the door to ultra-high-definition slow-motion footage. It also enables support for the Advanced Professional Video, or APV, codec, which Qualcomm says is aimed at pro-level recording workflows.

Audio has also been a focus. Both chips use AI to boost vocal clarity and reduce background noise, while Qualcomm’s new voice bubble technology is designed to isolate a user’s voice during calls. That could be useful in crowded public places, busy offices, or other noisy environments where call quality often suffers.

Camera and audio features at a glance

Why Qualcomm is betting on phones instead of dedicated AI devices

Qualcomm’s launch also reflects a broader industry belief that the phone remains the most important personal AI device, even as hardware companies explore standalone gadgets. The company has reportedly been working on more than 40 AI devices, but the central thesis behind its chip strategy is that most consumers will still reach for their phones first when they want AI assistance.

That view aligns with comments from several figures across the consumer tech world who believe the smartphone will remain the primary interface for personal AI. The logic is straightforward: people already carry their phones everywhere, unlock them dozens or hundreds of times a day, and use them for communication, photos, search, and productivity. Adding AI to that mix is a natural next step.

In that sense, Qualcomm’s new processors are less about a single product launch than about where the company believes the market is heading. The company is signaling to handset makers that the best way to compete in premium phones may be to build around local intelligence, not just benchmarks or camera specs.

Who is likely to use these chips first?

Premium Android phone makers are likely to be the first adopters, especially companies looking to market AI-heavy flagships. Qualcomm said Motorola announced its Motorola Signature 27 smartphone at the event, powered by the Snapdragon 8 Elite Extreme Gen 6, with general availability expected later this year.

That early announcement is important because it gives Qualcomm a concrete example of a phone maker already planning a device around the new silicon. It also suggests the company wants to move quickly from concept to commercialization, rather than waiting for a longer product cycle to make the AI case.

For smartphone brands, access to a chip like this could become a key differentiator. A handset that can run richer AI features locally may be easier to market around privacy, responsiveness, and premium imaging than one that depends heavily on cloud-based processing.

What does this mean for Android and the broader AI race?

It means the competition for AI leadership is moving deeper into hardware. The most visible AI products today are still software-first services and chatbots, but the next phase may be defined by which companies can put the best models into everyday devices without compromising battery life, heat, or responsiveness.

For Android, Qualcomm’s new chips could help ensure that flagship phones remain competitive in the AI era. They may also widen the gap between premium devices that can support advanced local models and lower-cost phones that must rely more heavily on cloud services.

For the broader industry, Qualcomm’s announcement reinforces a central theme: the future of consumer AI is increasingly about context. Devices that can hear better, remember more, and respond with less lag are likely to feel more useful than those that simply provide generic chatbot access.

Timeline of the Snapdragon 8 Elite Gen 6 launch

How the new chips fit Qualcomm’s larger strategy

Qualcomm’s message at the Snapdragon Summit was consistent: the company wants to be seen not just as a supplier of faster mobile processors, but as a platform provider for the AI phone era. That means making chips that can support assistants, memory, sensing, imaging, and voice processing all at once.

The move also reflects a competitive reality. Smartphone makers increasingly need a story that goes beyond incremental hardware gains. If AI is becoming the feature consumers care about most, then the chip inside the phone must be able to demonstrate tangible AI advantages, not just theoretical ones.

By highlighting local model support, voice features, and pro-grade video tools in the same announcement, Qualcomm is bundling the future of premium phones into one pitch: AI, creativity, and performance should all happen on the device, not somewhere else.

What comes next

The next test for Qualcomm will be how quickly phone makers translate this silicon into real products that consumers can buy and notice. A chip specification sheet can be impressive, but the market will judge the platform by whether the features feel genuinely helpful in day-to-day use.

If handset partners can deliver assistants that understand context, cameras that feel more flexible, and calls that sound cleaner in noisy environments, Qualcomm’s AI-first strategy could become a powerful differentiator. If not, the new chips may be seen as another generational boost with a heavy marketing layer.

Either way, the Snapdragon 8 Elite Gen 6 and Snapdragon 8 Elite Extreme Gen 6 show where Qualcomm thinks mobile computing is headed: toward phones that do not just run apps, but actively assist, observe, remember, and respond in real time.

Bottom line

Qualcomm’s latest flagship chips are a direct bet that smartphone buyers will value AI capability as much as speed, battery life, and camera quality. By pushing larger models, smarter sensing, and stronger media tools into its top-tier silicon, the company is trying to shape the next chapter of mobile computing around on-device intelligence.

• Qualcomm unveiled two new flagship chips focused on local AI performance.
• The Extreme version can run a 30-billion-parameter MoE model on-device.
• The chips add smarter sensing, camera improvements, and better voice processing.
• Motorola is among the first brands to announce a phone using the new silicon.
• The launch underscores how AI is becoming a central feature of premium smartphones.