Poolside Laguna XS.2
Hacker ApprovedRanked 29 of 39 open-weight language models by recommenders
Recommended for fast local use on CPU and for web research; critics say it trails rivals on benchmarks and is broken in places.
Recommended for fast local use on CPU and for web research; critics say it trails rivals on benchmarks and is broken in places.
15 people gave an opinion
- recommend it
- 12
- criticize it
- 3
- approval share
- 80%
Recommended in 6 of 13 months
- November 2025: not recommended
- December 2025: not recommended
- January 2026: not recommended
- February 2026: not recommended
- March 2026: not recommended
- April 2026: recommended
- May 2026: not recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: not recommended
- October 2026: recommended
- November 2026: recommended (month in progress)
Why it carries the badge
- 12 people recommend itNeeds 10 or more, met.
- 4 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 6 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Fast and good on CPU
“really fast on CPU and very good”
Comment from 27 Jun 2026 in “Ask HN: Running local LLMs? What's your model and hardware” “it's very good on cpu only”
Comment from 11 Jun 2026 in “Ask HN: What's the best LLM model that on a 24 GB VRAM GPU?” Agent harness works well
“theyre shipping the actual agent harness alongside the weights is the part that matters”
Comment from 28 Apr 2026 in “Laguna XS.2 and M.1” “It's fast, and the agent adheres to the ACP spec pretty well”
Comment from 28 Apr 2026 in “Laguna XS.2 and M.1” Competitive on coding benchmarks
“very nice to see it materialized into something competitive”
Comment from 28 Apr 2026 in “Laguna XS.2 and M.1” “reasonably competitive on coding related benchmarks”
Comment from 1 Jul 2026 in “Qwen 3.6 27B is the sweet spot for local development”
What people warn about
Each mark is one comment
Trails rivals on benchmarks
“not looking very good on the slightly-less-benchmaxxed Terminal-Bench 2.0”
Comment from 28 Apr 2026 in “Laguna XS.2 and M.1” “Not sure if this is competitive, look at the numbers for Qwen3.6”
Comment from 28 Apr 2026 in “Laguna XS.2 and M.1”
Other open-weight language models people recommend
The four most recommended other open-weight language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Qwen 3.8 27B | 1,173 | 426 | 73% | 13 | 2 mo32 owners | Approved |
| Z.ai GLM | 1,043 | 384 | 73% | 13 | 2 mo52 owners | Approved |
| Kimi | 931 | 359 | 72% | 13 | 1 mo26 owners | Approved |
| Gemma 4 | 713 | 249 | 74% | 13 | 1 mo14 owners | Approved |
| Poolside Laguna XS.2This page | 12 | 3 | 80% | 6 | — | Approved |
What people said
16 of the 19 comments that mention it, newest first. Each links to the original comment.
“I don't use it as much as Poolside's (excellent!!) Laguna XS 2.1 6bit”
“That is fast and almost as good.”
“Laguna drew random scribbles but the rest all managed to produce something monkey-like.”
“It was able to run locally with a context of 128k using llama-cpp and I tried using it for local development”
“without thinking enabled it is shockingly good at researching using web tool calls”
“reasonably competitive on coding related benchmarks”
“Oh! Looks like I’ve been sleeping on Laguna!”
“really fast on CPU and very good”
“it's very good on cpu only”
“It's broken, but it's very aesthetic”
“neither are particularly great”
“theyre shipping the actual agent harness alongside the weights is the part that matters”
“Not sure if this is competitive, look at the numbers for Qwen3.6”
“not looking very good on the slightly-less-benchmaxxed Terminal-Bench 2.0”
“very nice to see it materialized into something competitive”
“It's fast, and the agent adheres to the ACP spec pretty well”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.