Benchmark
Apex
Scores are percentages. Higher is better.
- Models measured
- 5
- Scores recorded
- 8
- Sources
- 2
- IQ weight
- 0.75
Score distribution
- 32.8%
- Mean
- 33%
- Median
- 23%
- Std dev
- 19.1%
- 10th percentile
- 44.5%
- 90th percentile
- 1%
- Min
This benchmark has a weight of 0.75 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Nemotron 3 Ultra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 84.8% |
| 2 | Qwen3.7 MaxQwen3.7-Max release · recommended reasoning-effort=xhigh system prompt· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 44.5% |
| 3 | Inkling 2 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 34.4% |
| 4 | DeepSeek V4 FlashDeepSeek V4 official model card · Think Max· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 33% |
| 5 | Qwen3.7 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 22.7% |
| 6 | Qwen3.7 PlusQwen3.7-Plus release · recommended reasoning-effort=xhigh system prompt· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 22.7% |
| 7 | DeepSeek V4 FlashDeepSeek V4 official model card · Think High· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 19.1% |
| 8 | DeepSeek V4 FlashDeepSeek V4 official model card · Non-Think· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 1% |
Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Data by LLM Stats