BALSAM - Fill in the Blank: 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) | 95.96 |
| 2 | GPT-5.2 | 85.61 |
| 3 | GLM-4.7 (Reasoning) | 82.88 |
| 4 | Claude Opus 4.5 | 81.87 |
| 5 | Grok 4.1 Fast (Reasoning) | 81.69 |
| 6 | GLM-5 (Thinking) | 79.58 |
| 7 | Llama 4 Maverick Instruct | 78.77 |
| 8 | Qwen 3 235B A22B | 78.75 |
| 9 | AceGPT-v2-8B-Chat | 77.64 |
| 10 | DeepSeek V3.2 | 77.2 |
| 11 | Kimi K2 Instruct (0905) | 76.08 |
| 12 | Command-R+ | 76.01 |
| 13 | Fanar-C-2-27B | 75.46 |
| 14 | Command A 03 2025 | 72.85 |
| 15 | Llama 3.3 70B Instruct | 72.04 |
Interactive version: theaggregate.ai/benchmark?slug=balsam-fill-in-the-blank · How It Works · Data refreshed daily, snapshot 2026-09-19.