Laguna XS.2
| Released | 2026-04-28 |
|---|---|
| Price per 1M | $0.000 in · $0.000 out |
| Overall score | 43.2 · rank 172 |
| Context window | 262,144 tokens |
| Provider | Poolside |
| Weights | Open weights |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 43.2 | 172 | 10 |
| Overall: Coding | 46.4 | 144 | 7 |
| Overall: Reasoning | 38.9 | 201 | 4 |
| Overall: Agentic work | 46.0 | 144 | 5 |
| Overall: Frontend | 42.2 | 164 | 5 |
| Coding | 50.0 | 117 | 3 |
| Agents | 46.0 | 130 | 1 |
| Reasoning | 39.2 | 201 | 1 |
| Math | 45.4 | 146 | 1 |
| Knowledge | 26.9 | 165 | 2 |
| Frontend | 38.2 | 115 | 2 |
Evidence
Coding 50.0 ±4.2
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SWE-bench Verified (Vals run) | 55.2% | Vals AI | 2026-09-01 |
| LiveCodeBench (Vals run) | 67.8% | Vals AI | 2026-09-01 |
| ProgramBench | 0.0% | Vals AI | 2026-11-09 |
Agents 46.0 ±3.4
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Terminal-Bench 2.1 (Vals run) | 25.8% | Vals AI | 2026-10-01 |
Reasoning 39.2 ±4.8
| Benchmark | Result | Source | Date |
|---|---|---|---|
| GPQA Diamond (Vals run) | 55.1% | Vals AI | 2026-09-01 |
Math 45.4 ±3.1
| Benchmark | Result | Source | Date |
|---|---|---|---|
| ProofBench | 0.0% | Epoch AI | 2026-11-11 |
Knowledge 26.9 ±3.9
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMLU-Pro (Vals run) | 69.1% | Vals AI | 2026-09-01 |
| Legal Research Bench | 1.0% | Vals AI | 2026-11-09 |
Frontend 38.2 ±0.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Vibe Code Bench v1.1 | 5.2% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,303 | LMArena | 2026-11-10 |
Names matched to this model
laguna-xs-2