Qualcomm’s new chip brings on-device AI to midrange phones

If you want to run generative AI directly on your smartphone, your biggest hurdle isn't tech—it's your wallet.

Most of the devices pushing on-device AI right now, like the Samsung Galaxy S24, start at $800 or more. Qualcomm wants to change that. On Tuesday, the chipmaker officially unveiled the Snapdragon 7s Gen 3, designed specifically to bring power-hungry generative AI features to phones that cost significantly less.

Flagship AI trickles down to $500

The Snapdragon 7s Gen 3 is aimed squarely at the $500 to $600 midrange smartphone market.

Qualcomm says the new silicon supports on-device generative AI capabilities, calling out features like real-time language translation and advanced noise cancellation. That means processing happens right on the hardware rather than relying on a cloud server.

For context, its predecessor, the 7s Gen 2, currently powers the $550 Motorola Edge. Devices sporting the new chip should land in that same ballpark.

Xiaomi will be the first brand to ship a phone with the 7s Gen 3, scheduled for an announcement next month. Qualcomm director of product management Akash Sharma also confirmed that Samsung, Sharp, and Realme plan to roll out devices powered by the chip in 2025.

A crowded midrange battlefield

Winning over phone makers in the midrange tier won't be a walk in the park for Qualcomm.

MediaTek is already putting up stiff competition with its Dimensity 7300 series, which powers devices like the $700 foldable Motorola Razr.

At the same time, phone manufacturers are leaning heavily into internal silicon. The $400 Samsung Galaxy A35 5G relies on Samsung's own Exynos 1380, while the $500 Google Pixel 8a runs on Google's flagship-grade Tensor G3.

Whether phone makers choose Qualcomm's new silicon ultimately hinges on their own margin and hardware strategies.

Why it matters

Generative AI on mobile has felt like a luxury reserved for those willing to drop nearly a grand on a flagship. Bringing those capabilities to sub-$600 devices is a huge win for everyday consumers who want local translation and smart audio processing without surrendering privacy or emptying their savings.

For the broader mobile ecosystem, this launch marks the transition of on-device AI from a novelty marketing perk to standard table stakes. Midrange handsets drive the bulk of global volume, and equipping them with AI-ready silicon forces competitors like MediaTek to follow suit.

That said, we think Qualcomm's ambition faces a tough reality check. Midrange phones frequently cut corners on RAM and thermal management, both of which are critical for smooth on-device model execution. Going toe-to-toe against Google's Tensor G3 in the $500 price bracket will be a steep climb unless OEM partners commit to matching software optimization.