Card: DeviceMark row (2026-09)
Browse files
README.md
CHANGED
|
@@ -24,6 +24,10 @@ base_model_relation: quantized
|
|
| 24 |
|
| 25 |
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).
|
| 26 |
|
|
|
|
|
|
|
|
|
|
|
|
|
| 27 |
# BitVLA (1.58-bit Vision-Language-Action) — Core AI
|
| 28 |
|
| 29 |
The **first Vision-Language-Action model running fully on-device on iPhone**, via Apple **Core AI**.
|
|
|
|
| 24 |
|
| 25 |
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).
|
| 26 |
|
| 27 |
+
<!-- gen-cards:devicemark begin (managed by scripts/gen-cards + tools/devicemark_row.py — edit cards.json, not this block) -->
|
| 28 |
+
This model has no row on [DeviceMark](https://devicemark.github.io/), the on-device LLM leaderboard.
|
| 29 |
+
<!-- gen-cards:devicemark end -->
|
| 30 |
+
|
| 31 |
# BitVLA (1.58-bit Vision-Language-Action) — Core AI
|
| 32 |
|
| 33 |
The **first Vision-Language-Action model running fully on-device on iPhone**, via Apple **Core AI**.
|