TestedActually.
Comparisons with published methodology Every claim dated & sourced

News · 2026-10-08

Microsoft optimises MAI-Code-1.1-Flash to run locally on Windows PCs with no per-call inference charge

At its 7 October 2026 Windows and Surface event Microsoft announced that MAI-Code-1.1-Flash now runs locally on Copilot+ PCs inside GitHub Copilot, with no inference charge for local model calls. The mixture-of-experts model has 137 billion total parameters with 6.8 billion active; the local build uses roughly 3-bit precision to cut size by about 80 percent to around 53GB while keeping a 256,000-token context window. Microsoft recommends more than 120GB of memory for best performance, and its Surface Laptop Ultra (from $2,599, shipping 16 October) carries up to 128GB of unified memory. The same event made Windows ML an experimental llama.cpp host, so GGUF models pulled from Hugging Face can run through new task-specific Windows APIs alongside ONNX.

What it means for buyers

Coding-assistant buyers gain a routed local option: Copilot can keep eligible coding work on-device at zero marginal inference cost once the hybrid routing preview lands later in October, which reframes the price-per-seat comparison in our coding-assistants table for anyone with Copilot+ hardware. For local-llm buyers, GGUF support inside Windows ML puts a first-party runtime beside Ollama and LM Studio, though the 53GB footprint and 120GB-plus memory guidance keep this a workstation lane for now rather than a mainstream laptop one.

Where to check it

Sources

  1. Windows Experience Blog: Building Windows for hybrid intelligence — vendor primary: hybrid intelligence strategy, local context, actions and models, checked 2026-10-08
  2. Microsoft AI: MAI-Code-1.1-Flash model page — vendor primary: local run in GitHub Copilot with no per-inference charges, checked 2026-10-08
  3. Microsoft Foundry on Windows: AI Development on Windows, from PyTorch and llama.cpp to Windows ML — vendor primary: experimental llama.cpp support and GGUF Text Generation API in Windows ML, checked 2026-10-08
  4. BizTech Weekly: Microsoft's Windows hybrid intelligence push — independent report: 137B total and 6.8B active parameters, 3-bit precision, 53GB, 120GB-plus memory guidance, Surface pricing, checked 2026-10-08