Apple adds Swift AI APIs, MLX upgrades and agent workflows across its platform stack

rxwei · x · 2026-07-25

Apple’s WWDC26 recap lays out five AI building blocks for Apple platforms:

• Foundation Models Framework: a native Swift API for on-device models, PCC, and third-party LLMs such as Gemini and Claude.
• Core AI: a framework for bringing custom models into Apple Silicon apps.
• MLX: open-source tooling for research, fine-tuning, local inference servers, and distributed computing.
• Agentic experiences: multi-model workflows with context handoffs between specialized agents.
• Evaluations: automated quality and failure-detection tools to help teams ship with confidence.

Original post →

More from Infra

• SemiAnalysis says AMD still faces ROCm instability and MI455X ramp issues — zephyr_z9 · 2026-07-25
• Intel consumer motherboards can break PCIe P2P on multi-GPU AI rigs — Arli_AI · 2026-07-25
• Scale-out networking doesn’t need CPO, slide argues; compute still dominates power use — zephyr_z9 · 2026-07-25
• Open-source DKV framework cuts KV-cache memory for long-context local inference — Om_5000 · 2026-07-25
• Kimi weights could turn the debate into hardware economics versus V4 — teortaxesTex · 2026-07-25
• Jensen Huang hands Elon Musk a desk-size DGX Spark at Starbase — XFreeze · 2026-07-25

Infra channel →