Skip to content

Benchmark

RULER 32K

Long context benchmarks measure how well models use information in long inputs. Scores are percentages. Higher is better.

Models measured
28
Scores recorded
28
Sources
2
IQ weight
No weight published

Score distribution

87.9%
Mean
92.4%
Median
10.2%
Std dev
69%
10th percentile
96%
90th percentile
63%
Min
63%97.3%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1Gemma 4 26B A4BGemma 4 Technical Report · Table 9 · without thinking · accuracy · 32K· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.97.3%
2Gemma 4 31B ITGemma 4 Technical Report · Table 9 · without thinking · accuracy · 32K· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.96.8%

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

ModelScore
C4AI Command R 08 2024Command-R-0824, 32B· RULER official results data90.8%
DBRX-instructDBRX, 36B/132B· RULER official results data63.1%
Gemini 1.5 ProGemini-1.5-pro· RULER official results data95.9%
GPT 4 1106 PreviewGPT-4-1106-preview· RULER official results data93.2%
Jamba 1.5 LargeJamba-1.5-large, 94B/398B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.96%
Jamba 1.5 MiniJamba-1.5-mini, 12B/52B· RULER official results data94.6%
Llama 3.1 70B InstructLlama3.1, 70B· RULER official results data94.8%
Llama 3.1 8B InstructLlama3.1, 8B· RULER official results data87.4%
Llama-2-7B-32K-InstructTogether, 7B· RULER official results data63%
Mistral 7B Instruct v0.2Mistral, 7B· RULER official results data75.4%
Mistral Large Instruct 2411Mistral-Large-2411, 123B· RULER official results data94%
Mistral NemoMistral-Nemo· RULER official results data69%
Mistral-Large-Instruct-2407Mistral-Large-2407, 123B· RULER official results data93%
Mixtral 8x22B Instruct v0.1Mixtral-8x22B, 39B/141B· RULER official results data90.9%
Mixtral 8x7B Instruct v0.1Mixtral-8x7B, 12.9B/46.7B· RULER official results data85.9%
Phi-3 Medium 128K InstructPhi3-medium, 14B· RULER official results data86.8%
Phi-3 Mini 128K InstructPhi3-mini, 3.8B· RULER official results data87.5%
Qwen 3 235B A22BQwen3-235B-A22B, 235B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.95.1%
Qwen1.5-72B-ChatQwen1.5, 72B· RULER official results data67.8%
Qwen2 72B InstructQwen2, 72B· RULER official results data94.1%
Qwen3 14BQwen3-14B, 14B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.96.1%
Qwen3 30B A3BQwen3-30B-A3B, 30B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.92.4%
Qwen3 32BQwen3-32B, 32B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.94.4%
Qwen3 4BQwen3-4B, 4B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.87.8%
Qwen3 8BQwen3-8B, 8B· RULER official results data· Reported by the model's developer. The source relaying it did not verify it.91.2%
Yi-34BYi, 34B, 200k· RULER official results data87.5%

Data credits: NVIDIA RULER, Hsieh et al. Official README numeric table; no separate results licence asserted. · 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.