Laguna S 2.1 (118B-A8B, Q6 MLX)
Poolside's Laguna S 2.1 in a calibrated oQ6e MLX quantization for native Apple Silicon inference. The model has 118B total parameters with about 8B active per token and is built for agentic coding, long-horizon software work, native tool calling, and interleaved reasoning. This MLX-only build is about 92GB on disk, measured about 90GB peak at 64K context, and targets 128GB Macs. Gezel defaults to Poolside's quality-recommended 256K context window; the underlying checkpoint can be extended to 1M context with explicit YaRN overrides.
At a glance
- Parameters
- 118B
- Quantization
- oQ6e
- Context window
- 262K tokens
- Approx. size
- 92.0 GB
- Engines
- MLX
- License
- OpenMDW-1.1 (open)
- Version
- v1.0.0 , released 2026-07-22
- Family
- glm
Tuning
Defaults a gezel client applies out of the box, from real evaluation runs against this exact quantization.
- sampling.temperature
0.7- sampling.topP
0.95- sampling.maxTokens
8192- reasoning.thinkingBudget
4096- reasoning.enableThinking
true- profiles
- thinking-general, thinking-coding, thinking-precise, instruct, creative
Model behaviors
Client-side behavior modules the catalog enables for this model (reasoning-tag handling, fabrication detection, prompt shaping).
- reasoning.strip-think-tags
- prompt.private-reasoning-guidance
- fabrication.detect-past-tense-no-tools
- turn.preamble-folding
- turn.ramble-detection
- tools.mlx-grammar
Sources
- MLX
- mlx-community/Laguna-S-2.1-oQ6e
- Upstream
- https://huggingface.co/poolside/Laguna-S-2.1
- Manifest
- View on GitHub
Tags
- poolside
- coding
- agentic
- tools
- reasoning
- long-context
- moe
- large
- open-weights
- apple-silicon