Card: definition line (Core AI, 2026-09)

README.md

@@ -12,6 +12,8 @@ tags: [core-ai, coreaikit, asr, speech-to-text, text-normalization,
        post-processing, on-device, apple, qwen3]
 ---
 
+Core AI is Apple's on-device ML runtime in iOS 27 / macOS 27 and the successor to Core ML: PyTorch models are exported with Apple's `coreai-torch` (LLMs: `coreai.llm.export`) into `.aimodel` bundles that run on the GPU or the Neural Engine, e.g. Qwen3-8B 4-bit decodes at 94 tok/s on an M4 Max GPU, MLX 90 under the same protocol ([apple-silicon-llm-bench](https://github.com/john-rocky/apple-silicon-llm-bench), macOS 27 beta, 2026-06).
+
 # S1-mini by Superwhisper — Core AI
 
 [**S1-mini**](https://huggingface.co/superwhisper/s1-mini) by **Superwhisper** converted to