Skip to main content
AI Choice Engine

Poolside

Poolside Laguna M.1

Poolside Laguna M.1 is a Poolside model with 262K context, Not verified in / Not verified out per million tokens, last verified October 25, 2026. Open weights: yes. Public-board scores are linked to the publisher; vendor-only figures stay in the watch-out, not on the board.

Laguna M.1 is Poolside's open-weight coding model with a 256K context.

Recorded profile verification October 25, 2026Individual provider fields may change

AI model specification details
SpecificationPoolside Laguna M.1
ProviderPoolside
TierBalanced
Context window262K
Max outputNot verifiedUnverified
Input / 1M tokensNot verifiedUnverified
Output / 1M tokensNot verifiedUnverified
WeightsOpen
Parameters225B total / 23B active (MoE)
Reasoning levelsNot verifiedUnverified
Modalitiestext
LicenseApache 2.0
ReleasedApril 28, 2026

Pricing tiers: Apache 2.0 downloadable 225B/23B MoE with self-host compute costs and a 262,144-position checkpoint config. April 28 hosted access was free for a limited time; the current OpenRouter M.1 entry has no endpoints, so no current hosted rate or availability is asserted. BF16, FP8 and NVFP4 weights exist; their repository timestamps do not prove the public weights release day.

Best for

  • Open-weight coding
  • Repo-scale work
  • Self-hosting

Watch out

Thinking is enabled by default in the documented setup and can be disabled per request. Preserve reasoning_content in follow-up assistant messages; dropping it can prevent later reasoning. A 225B checkpoint requires substantial memory. The vendor card used pool harness, 500 steps, temperature=1.0/top_k=20, 256K context, four runs per task and patched task images; these figures are not independent public-board measurements. A model-specific hosted hard output cap remains unverified.

Source receipts

Catalog figures for this model were checked against the following sources.

Common questions

Poolside Laguna M.1

Answered from the verified figures on this page rather than general guidance.

What is Poolside Laguna M.1's context window?
Poolside Laguna M.1 accepts about 262K tokens of context. That only matters if you routinely send very long documents, large codebases, or multi-turn histories that approach that limit.
What is Poolside Laguna M.1 best for?
Poolside Laguna M.1 is a balanced tier from Poolside. It suits open-weight coding, repo-scale work, self-hosting. Thinking is enabled by default in the documented setup and can be disabled per request. Preserve reasoning_content in follow-up assistant messages; dropping it can prevent later reasoning. A 225B checkpoint requires substantial memory. The vendor card used pool harness, 500 steps, temperature=1.0/top_k=20, 256K context, four runs per task and patched task images; these figures are not independent public-board measurements. A model-specific hosted hard output cap remains unverified.
Can I self-host Poolside Laguna M.1?
Poolside Laguna M.1 publishes open weights, but self-hosting depends on the licence, hardware footprint, quantisation quality, and serving stack. A hosted API is often cheaper until you have measured throughput and concurrency on your own hardware.
Model comparison catalogBenchmark library