Laguna XS.2 — benchmark results
Poolside's open agentic-coding MoE (33B total, 3B active) for single-GPU local use. Provider: Poolside. Released 2026-04-28. Access: Open.
Unified ELO 1544 ± 36, rank #641 of 1776 rated models, from 22 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| SvelteBench | 92.2 | Average pass@1 (%) | 71.6 |
| Wolfram LLM Benchmarking Project | 39 | Correct Functionality (%) | 47.4 |
| Vals AI LiveCodeBench | 66.61 | Accuracy (%) | 36.2 |
| Vals AI CorpFin v2 | 56.33 | Accuracy (%) | 34.4 |
| Vals AI Terminal-Bench 2.0 | 28.09 | Accuracy (%) | 28.8 |
| Vals AI LegalBench | 71.03 | Accuracy (%) | 20.8 |
| Chatbot Arena (Code) | 1303 | Elo | 20.2 |
| PM-LLM-Benchmark | 25.6 | Score | 18.9 |
| Vals AI GPQA | 58.08 | Accuracy (%) | 18 |
| Vals AI Vibe Code Bench | 5.21 | Accuracy (%) | 14.3 |
| Vals AI ProgramBench | 16.5 | Raw Pass Rate (%) | 13.3 |
| Vals AI MMLU-Pro | 69.41 | Accuracy (%) | 10.7 |
Interactive version: theaggregate.ai/model?slug=laguna-xs-2 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.