Poolside drops Laguna XS.2 open weights

// 45d agoMODEL RELEASE

Poolside drops Laguna XS.2 open weights

Poolside has released Laguna XS.2, a 33B MoE model with 3B active parameters, as its first open-weight model under Apache 2.0. It is aimed at long-horizon agentic coding workflows and is positioned as a compact, locally runnable option with strong benchmark results for its size.

// ANALYSIS

Hot take: this is a meaningful open-weight coding release, not just another small model drop. Poolside is clearly aiming at the “serious agentic coding” segment, where long context, tool use, and local deployability matter more than raw parameter count.

–33B total parameters with 3B active parameters per token, built as a Mixture-of-Experts model.
–Released under Apache 2.0, which makes it unusually usable for commercial and community reuse.
–Positioned for long-horizon coding tasks and agent workflows, not generic chat.
–Poolside is also surfacing benchmark claims that put it in a credible small-model performance tier.
–The release is paired with broader ecosystem support, including local deployment and framework integration.

// TAGS

llmagentopen-sourcedevtoolbenchmark

DISCOVERED

45d ago

2026-04-29

PUBLISHED

45d ago

2026-04-29

RELEVANCE

9/ 10

AUTHOR

WorldofAI

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE24m ago

Grok Build worktrees spark developer praise

A developer on X praises the fast-paced development of Grok Build, an AI coding tool from xAI, noting that the team has shipped numerous updates recently. They specifically recommend that developers explore the new worktrees feature over the weekend to experience the product's latest capabilities.

OPEN SOURCE37m ago

Mosh is an open-source, model-driven application security testing harness that wraps around LLMs to automate penetration testing through discovery, planning, dockerized execution, and reporting.

Mosh (Model-driven Open Security Harness) is an open-source security testing application designed to automate the work of a security researcher. Instead of relying on raw prompts, the tool implements a multi-step workflow starting with application discovery (mapping routes and technologies), security planning (creating test hypotheses), and controlled test execution through Docker containers using engagement settings. It continuously writes structured reports and memory logs, allowing developers to safely run, review, and reproduce pen-testing results iteratively as vulnerabilities are resolved.

OPEN SOURCE41m ago

Zhipu AI to open-source GLM-5.2 under MIT

Zhipu AI has announced plans to open-source its flagship GLM-5.2 coding model under the permissive MIT license next week. The model features a 1-million-token context window and is currently deployed on Zhipu's GLM Coding Plan.

Poolside drops Laguna XS.2 open weights