internlm2.5-20B-chat: benchmark results
Shanghai AI Lab's open 20B InternLM2.5 chat model (August 2024), reasoning-focused with long-context and tool/web-search support. Provider: Shanghai AI Lab. Released 2024-07-30. Access: Open.
Aggregate Rating 1497 ± 26, rank #799 of 1661 rated models, from 44 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open LLM Leaderboard - BBH | 74.74 | Score | 99.9 |
| Open LLM Leaderboard - MATH Level 5 | 40.79 | Score | 90.6 |
| Open LLM Leaderboard - MuSR | 45.58 | Score | 86 |
| Open LLM Leaderboard - IFEval | 70.1 | Score | 83.1 |
| Open Portuguese LLM - BLUEX | 61.34 | Accuracy (%) | 80.2 |
| Open LLM Leaderboard - GPQA | 32.13 | Score | 76.1 |
| Open LLM Leaderboard - MMLU-Pro | 39.98 | Score | 73.7 |
| FinEval - Financial Academic Knowledge (Zero-shot CoT) | 67.3 | Accuracy (%, zero-shot chain-of-thought) | 72.2 |
| FinEval - Financial Industry Knowledge | 53.2 | Similarity score (%, zero-shot) | 66.7 |
| Open Portuguese LLM - ENEM | 67.32 | Accuracy (%) | 66.1 |
| Open Portuguese LLM - OAB Exams | 45.42 | Accuracy (%) | 60.8 |
| FinEval - Financial Security Knowledge | 74.1 | Accuracy (%, zero-shot) | 55.6 |
Interactive version: theaggregate.ai/model?slug=internlm2-5-20b-chat · How It Works · Data refreshed daily, snapshot 2026-10-23.