BALSAM - Summarization: leaderboard
Metric: Overall score (0-100, LLM-judged generation and multiple choice). Source: benchmarks.ksaa.gov.sa. 29 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemma 4 31B (IT) | 69.18 |
| 2 | DeepSeek V4 Pro (Reasoning) | 62.77 |
| 3 | GLM-5.2 | 55.7 |
| 4 | Qwen 3 235B A22B | 52.64 |
| 5 | GLM-4.7 (Reasoning) | 52.09 |
| 6 | GLM-5 (Thinking) | 51.9 |
| 7 | GPT-5.2 | 51.55 |
| 8 | Kimi K2 Instruct (0905) | 51.44 |
| 9 | DeepSeek V3.2 | 50.46 |
| 10 | AceGPT-v2-8B-Chat | 50.25 |
| 11 | Claude Opus 4.5 | 49.86 |
| 12 | Grok 4.1 Fast (Reasoning) | 49.71 |
| 13 | Command A 03 2025 | 49.35 |
| 14 | Fanar-C-2-27B | 49.17 |
| 15 | MiniMax M3 (Reasoning) | 49 |
Interactive version: theaggregate.ai/benchmark?slug=balsam-summarization · How It Works · Data refreshed daily, snapshot 2026-09-19.