TensorOps is a Metal Shading Language library that lets developers write optimized custom machine learning kernels—including matrix multiplication, convolution, and FlashAttention—with automatic hardware acceleration across all Apple Silicon GPU generations, including the new M5 neural accelerator. iOS/macOS 27 extends its quantized data type support to FP8, 2-bit integers, and MX scaling formats.
• Added FP8 (E4M3, E5M2) and 2-bit integer quantized data types on top of the 4/8-bit integers added in iOS 26
• Single MTLTensor can now carry an auxiliary scale plane in FP8 E8M0 block-wise format alongside quantized data
• Cooperative tensors can now be fed directly as inputs to matmul2d ops without a threadgroup memory store/reload step
• Full neural accelerator utilization on M5 via TensorOps without any code changes
• Enables plug-in custom ML operations for high-level frameworks like Core AI and MLX without sacrificing GPU performance
• New FP8/2-bit quantization and E8M0 block-wise scale factors reduce memory bandwidth and let larger models fit on device
• Cooperative tensors can now be passed directly as matmul inputs, eliminating costly threadgroup memory round-trips in fused kernels like FlashAttention
Foundation Models is a new Apple framework introduced in iOS 27 that gives developers on-device access to the same Apple Intelligence language model powering system features, enabling text generation, structured output, and tool-calling entirely on-device without a network connection.
Visual Intelligence brings iOS 17's Visual Look Up capabilities to a new developer-facing API surface in iOS 27, letting apps pipe live camera frames or static images through on-device scene understanding to extract subjects, text, barcodes, and rich semantic labels without any cloud round-trip.
iOS 27 opens the Foundation Models framework to third-party LLM providers via a new public LanguageModel protocol, enabling anyone to integrate custom, server-based, or open-source models using the same Swift API as Apple's on-device system model.
In-depth guide
iOS 27 On-Device AI & Apple Intelligence →App Schemas let developers describe their app's content and actions using pre-defined domain schemas (like the Calendar domain) so Siri can understand, search, and act on app data without custom NLP. Entities conforming to IndexedEntity are donated to Spotlight's semantic index, enabling natural-language queries over app content.