Apple NASDAQ:AAPL may have a new path for on-device AI after startup PrismML reportedly shrank Alibaba's NYSE:BABA Qwen 3.6 model to run on the iPhone 17 Pro.
The Information reported that PrismML compressed the 27 billion-parameter model to less than 4GB from about 54GB, while claiming the reduction does not hurt performance. The model can reportedly handle complex chat, reasoning, autonomous agents and software coding, with all 27 billion parameters active at once.
Apple has held meetings with PrismML about possible uses of its technology, according to the report. The startup's approach could appeal to Apple, which has emphasized running more AI features on devices rather than relying fully on cloud data centers.
the report highlights Apple's push to close the AI gap without matching the data-center spending arms race led by Microsoft NASDAQ:MSFT, Amazon NASDAQ:AMZN, Meta NASDAQ:META and Google NASDAQ:GOOG. The next test is whether PrismML's compression method can scale beyond a technical milestone into real Apple product features.