Apple is exploring new ways to keep AI processing on users’ devices, prioritizing speed, privacy, and reduced reliance on the cloud. The company is in early discussions with PrismML, a startup developing technology that dramatically compresses large AI models to fit and run efficiently on iPhones.

According to PrismML CEO Babak Hassibi, who spoke to CNBC, Apple and other companies have been evaluating the startup’s models, measuring their speed, energy efficiency, and performance on devices. “They’re really evaluating our technology right now,” Hassibi said of Apple. He described the discussions as very early but noted that “things are progressing nicely.”

Shrinking Massive Models for Everyday Devices

PrismML, a Khosla Ventures-backed spinout from the California Institute of Technology, recently released compressed versions of Alibaba’s open-source Qwen model. The company reduced the model from roughly 54 GB to less than 4 GB, enabling all 27 billion parameters to run on an iPhone 15 or newer.

The compression technique simplifies how internal information is stored—reducing values from 16 bits to just one or three possible values. PrismML claims the compressed models use 10 to 15 times less memory, generate responses six to eight times faster, and consume three to six times less energy than conventional versions.

Hassibi acknowledged a small performance trade-off: “PrismML’s models typically lose a few percentage points of overall performance, with factual recall weakening before skills such as reasoning, math and coding.” Despite this, the company is releasing two compressed versions for free, designed to run on iPhones, MacBooks, and Nvidia-powered PCs.

Boosting Siri and On-Device Capabilities

This development aligns closely with Apple’s ongoing efforts to enhance Siri. Just before PrismML’s release, Apple opened the public beta of iOS 27, introducing a major overhaul aimed at making Siri more competitive with offerings from xAI, OpenAI, and Anthropic — while emphasizing on-device processing for personal data.

Running more AI locally would bring several benefits:

• Reduced latency compared to cloud requests
• Lower cloud computing costs
• Support for features that work offline
• Stronger privacy protections, especially for sensitive health, medication, or personal information

Carolina Milanesi, president and principal analyst at Creative Strategies, told CNBC: “The more you can do on device, the better it is,” highlighting applications like computational photography, video generation, and health tools.

Analysts note Apple’s integrated hardware-software approach gives it an edge in optimizing these models. Horace Dediu of Asymco explained that Apple aims to keep most common Siri interactions on-device while routing only the most demanding tasks to the cloud. “They’re trying to figure out how big a model and how clever a model they can fit on the device,” he said.

Potential Industry Impact and Caveats

If PrismML’s technology proves effective at scale, it could influence memory requirements and datacenter demands. However, analysts caution that real-world performance — especially on lengthy prompts, multitasking, battery life, and millions of queries — remains to be fully validated.

Tarun Pathak of Counterpoint Research emphasized: “The ultimate test will be millions of queries, thousands of device combinations and robust testing at scale.” Phil Solis of IDC pointed to power consumption as a key open question for always-on or agent-like features.

Hassibi sees broader applications beyond phones, including robotics and autonomous systems where fast, local intelligence is critical: “It’s very important that the intelligence be local and that it can run fast.”

Pathak summed up the balanced future: “The combination of cloud and on-device AI can serve a more complete, efficient and privacy-centric AI experience. Complex tasks will be offloaded to the cloud, whereas sensitive, latency-critical and privacy-relevant tasks will be executed on-device.”

MacDailyNews Take: Apple’s interest in technologies like PrismML underscores its commitment to privacy as a core differentiator while pushing the boundaries of what smartphones can achieve with AI. As testing continues, this could mark a significant step toward more capable, personal, and private intelligent assistants.

‎
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!

Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.

The post Privacy-focused Apple in talks with startup to run more powerful AI models directly on devices appeared first on MacDailyNews.

Apple is in discussions with PrismML, a Khosla Ventures-backed startup that claims to have achieved a major breakthrough in running exceptionally large AI models directly on iPhones without relying on cloud servers, Aaron Tilley reports for The Information. The company says it has successfully compressed and executed a 27-billion-parameter model on the latest iPhone hardware, marking what it describes as the largest AI model ever run fully on-device.

PrismML, which emerged from stealth earlier this year as a Caltech spinout, specializes in extreme model compression techniques, particularly 1-bit and ternary weight architectures. These methods drastically reduce memory footprints while aiming to preserve performance. According to the startup, its technology allowed Alibaba’s open-source Qwen 3.6 (27B parameters) — which normally requires about 54 GB in standard precision — to shrink to under 4 GB. The compressed model reportedly ran with all parameters active simultaneously on an iPhone 17 Pro, enabling more complex on-device tasks like advanced reasoning and software development assistance.

This development aligns with Apple’s long-standing push to enhance Apple Intelligence with more powerful on-device capabilities. By keeping AI processing local, Apple can improve privacy, reduce latency, and lessen dependence on cloud infrastructure and external partners. Sources familiar with the matter indicate Apple has held talks with PrismML about potential integration of its compression technology to expand what future iPhones can handle natively.

The Technical Leap: 1-Bit Models and Intelligence Density

Traditional large language models use 16-bit floating-point precision, leading to massive memory requirements. PrismML’s approach uses 1-bit weights (essentially on/off states) and ternary variants (weights of -1, 0, or +1). The company’s flagship open-source “Bonsai” family includes:

• Bonsai 8B: An 8-billion-parameter model that fits in roughly 1 GB (versus ~16 GB for a typical 16-bit equivalent), while delivering competitive results on benchmarks like MMLU, GSM8K, and HumanEval.

• Smaller variants (4B and 1.7B) optimized for even greater efficiency.

PrismML claims these models offer up to 14× smaller memory use, 8× faster inference, and 5× better energy efficiency compared to full-precision counterparts, resulting in over 10× “intelligence density.” The startup has open-sourced its models under Apache 2.0 and released custom kernels for Apple’s Metal framework, making them compatible with iPhone and Mac hardware.

Co-founders, including Caltech professor Babak Hassibi and other PhDs, emphasize that the breakthrough stems from years of research into neural network compression. Investors include Khosla Ventures, Cerberus Capital, and Caltech itself, with the company raising $16.25 million in seed funding. Vinod Khosla has praised the work as a “mathematical breakthrough” that could shift AI from data-center dominance to efficient edge deployment.

Implications for Apple and the Industry

For Apple, adopting such technology could supercharge features in iOS and Apple Intelligence, enabling more sophisticated AI agents, real-time coding help, or advanced multimodal capabilities without constant server calls. It also addresses growing concerns around energy consumption and data privacy in AI.

PrismML positions its tech not just for consumer devices but for robotics, wearables, and industrial edge applications. By open-sourcing core models, the company hopes to accelerate adoption and invite community improvements.

While independent verification of all performance claims is still emerging, the reported iPhone demo and Apple’s interest signal strong validation. As the AI race intensifies, solutions that pack more intelligence into everyday devices — without ballooning costs or power draw — could prove transformative.

PrismML plans further releases and is expected to continue engaging with major device makers.

MacDailyNews Take: For Apple, which has invested heavily in on-device AI silicon like the Neural Engine, this could represent a key for the next-gen of truly personal, secure, and private artificial intelligence.

‎
Please help support MacDailyNews — and enjoy subscriber-only articles, comments, chat, and more — by subscribing to our Substack: macdailynews.substack.com. Thank you!

Support MacDailyNews at no extra cost to you by using this link to shop at Amazon.

The post Apple eyes PrismML to run huge AI models directly on iPhone appeared first on MacDailyNews.