100-Billion Parameter AI Models to Run on Phones by 2028
Developers aim to enable private on-device agents by 2028, leveraging personal sensor data without reliance on cloud processing.
Updated on Oct. 16, 2026 in Consumer Electronics
Live Poll
Would you pay higher prices for a smartphone to run advanced AI models locally by 2028?
AI developers are pushing to run 100-billion parameter models locally on smartphones by 2028. This move aims to leverage device-level sensor data and personal context for more capable AI agents.
Why it matters
Running massive models on-device allows AI to access private data without cloud latency, but current hardware constraints and memory costs have slowed the smartphone market by 20 percent this year.
Current memory prices have increased five to six times relative to previous baselines, impacting the scalability of the 1.2 billion annual smartphone market.
The players
Qualcomm
A semiconductor company that produces processors for mobile and smart devices and is developing 6G and AI platform standards.
The details
Running models on smartphones requires local access to sensor data, personal context, and installed applications, allowing agents to function without constant cloud queries. Qualcomm plans to open-source its Modular platform under the Apache 2.0 license to help developers port models across varying chip architectures. This push for on-device processing relies on future hardware efficiency gains to manage the complexity of 100-billion parameter systems.
Timeline
2028: Targeted deadline for running 100-billion parameter AI models on smartphones.
2028: Scheduled 6G network demonstrations at the Los Angeles Olympics.
2029-2030: Anticipated window for early 6G network deployments.
The Tech Race
The industry is using the 2028 Los Angeles Olympics as a milestone for demonstrating mobile AI and 6G integration. This timeline establishes a clear competitive roadmap for hardware developers to shift from cloud-dependent models to local, high-parameter AI execution.
Users can expect AI agents to become more personalized as models gain direct access to local sensor and app data. However, the current memory price spikes suggest that mobile hardware costs will remain high even after supply shortages stabilize.
The takeaway
Developers aim to replace cloud-dependent AI with 100-billion parameter local models by 2028 to improve privacy and responsiveness. Watch for 6G deployment timelines around the 2029-2030 period as the necessary bandwidth for these systems matures.
Further reading
Explore how shifting component standards are influencing device capabilities in our Consumer Electronics section.
Source note: This article includes information reported by The Next Web.
Live Poll
Would you pay higher prices for a smartphone to run advanced AI models locally by 2028?
