Laguna XS 2.1.

A compact-looking deep-water coder: 33B total parameters, only 3B active per token, and enough patience for the work that unfolds over many steps.

View on Hugging Face See LocalEngine variants ↓

Model makerPoolside

Supported in LocalEnginemacOS

Upstream licenseOpenMDW 1.1

SourceLaguna XS 2.1 model card ↗

Where it earns its space

Built for a real local rhythm.

These are good-fit starting points—not promises of benchmark performance. Choose the smallest listed option that comfortably covers your work.

Long-horizon coding tasks

Agentic software workflows

Reasoning over a larger local codebase

In LocalEngine

Configured variants, by device.

The formats below are the downloads currently represented in LocalEngine’s model catalogs. GGUF runs through the llama.cpp-based runtime; MLX options use the Apple Silicon MLX runtime.

macOS

33B total · 3B active

Formats: GGUF Q4_K_M · BF16
Local note: Q4_K_M is the practical local format and needs at least 36 GB unified memory; BF16 needs at least 72 GB.

About the source

Model identity, task category, supported formats, and license are based on the linked Hugging Face Laguna XS 2.1 model card and the LocalEngine macOS/iOS catalogs. Hardware guidance reflects the packaged variants, not a universal system requirement.