Skip to content

Benchmark

LifeSciBench

Scores are percentages. Higher is better.

Models measured
7
Scores recorded
12
Sources
2
IQ weight
0.10

Score distribution

56.4%
Mean
57.6%
Median
3.7%
Std dev
51.2%
10th percentile
60.3%
90th percentile
50.4%
Min
50.4%60.3%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.10 for chat models, measured over 7 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1GPT-6 Astra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim60.3%
2GPT-6 AstraOpenAI launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.60.3%
3GPT-5.6 Sol· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim59.9%
4GPT-5.6 SolOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.59.9%
5GPT-5.6 SolOpenAI launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.59.9%
6GPT-rosalind-research· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim57.6%
7GPT-5.6 Terra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim56%
8GPT-5.6 TerraOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.56%
9Claude Opus 4.8OpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.53.6%
10GPT-5.6 Luna· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim51.2%
11GPT-5.6 LunaOpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.51.2%
12GPT 5.5OpenAI GPT-5.6 launch table· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.50.4%

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