Qwen2.5-1.5B-Instruct
Qwen/Qwen2.5-1.5B-Instruct
Qwen · Language models · family qwen2.5
macOSiOS
Apple’s model card
Apple lists this model in its Core AI model registry, which means there is a supported recipe for converting it to the .aimodel format. The converted model is not hosted by Apple; you run the export yourself with the command below.
Export presets
| Preset | Platform | Compression | Precision | Context |
|---|---|---|---|---|
| qwen2.5-1.5b-instruct | macOS | 4bit | float16 | 32K |
| qwen2.5-1.5b-instruct-8bit-kv | macOS | 4bit_weights_8bit_kv_cache | float16 | 32K |
| INT4 per-block weights and INT8 per-tensor KV cache. | ||||
| qwen2.5-1.5b-instruct | iOS | 4bit_weight_palettized_group8 | float16 | 4K |
Export it
From a checkout of apple/coreai-models, with uv installed:
uv run coreai.llm.export qwen2.5-1.5b-instruct uv run coreai.llm.export qwen2.5-1.5b-instruct-8bit-kv uv run coreai.llm.export qwen2.5-1.5b-instruct --platform iOS
Links
First seen in the catalogue 2026-11-11.