PaLM 2: benchmark results
Provider: Google. Released 2023-05-10. Access: API.
Aggregate Rating 1423 ± 25, rank #1117 of 1640 rated models, from 40 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| NPHardEval - GCP | 16.18 | Accuracy (%) | 90.9 |
| NPHardEval - KSP | 2 | Accuracy (%) | 90.9 |
| NPHardEval - TSP D | 56.55 | Accuracy (%) | 72.7 |
| MoralChoice | 99.3 | MoralChoice Acc (%) | 72.5 |
| CodeScope - Code Repair | 7.24 | Score | 71.4 |
| NPHardEval - SPP | 21.82 | Accuracy (%) | 63.6 |
| NPHardEval - SAS | 41.64 | Accuracy (%) | 45.5 |
| NPHardEval - TSP | 0.73 | Accuracy (%) | 45.5 |
| CodeScope | 26.37 | Score | 42.9 |
| CodeScope - Code Generation | 5.46 | Score | 42.9 |
| CodeScope - Code Translation | 3.51 | Score | 42.9 |
| CodeScope - Code Understanding | 47.28 | Score | 42.9 |
Interactive version: theaggregate.ai/model?slug=palm-2 · How It Works · Data refreshed daily, snapshot 2026-11-07.