Building local AI systems that bring large open models to laptop CPUs.
专注于让大模型在普通笔记本 CPU 上真正可用。
Native C · CPU inference · low-memory runtimes · reproducible performance
Run the best model under 200B on a single laptop CPU with a purpose-built native C inference engine. Its 8 GB RAM path is selected automatically; no GPU, Python, PyTorch, model conversion or external inference runtime is required. Measured performance reaches 9.89 positions/s for exact batch-4 verification and 5.03 token/s in resident single-conversation chat. Runtime optimizations add no approximation beyond the selected quantized GGUF.
Run the native 284B-A13B DeepSeek-V4-Flash-0731 checkpoint on one laptop CPU. The pure C engine streams cold MoE experts from disk, has a tested 8 GB RAM path, needs no GPU or Python, and reaches 1.12 token/s in its best documented prompt-lookup workload. It supports terminal chat and a resident local OpenAI-compatible API with live token streaming, parallel function calls, and matched tool-result replay.
Run Qwen3.8-27B locally on one laptop CPU with a native C engine, direct GGUF loading, an 8 GB RAM tested path and measured generation up to 2.52 token/s. Chat in the terminal or connect local apps through its resident OpenAI-compatible API with parallel function calls and tool-result replay. Runtime speedups preserve byte-identical full logits against the native baseline for the same GGUF.
- Native inference engines instead of wrappers around external runtimes
- Low-bit CPU kernels, bounded memory planning and storage-aware execution
- Reproducible TTFT, TPOT, memory and correctness evidence
- User paths that go from clone to a real conversation without a GPU