Skip to content

Benchmark

GPQA

Science benchmarks measure how well models answer science questions. Scores are percentages. Higher is better.

Models and systems measured
364
Scores recorded
378
Sources
3
IQ weight
0.75

Score distribution

45.3%
Mean
47.5%
Median
34.4%
Std dev
2.9%
10th percentile
89%
90th percentile
0%
Min
0%97%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.75 for chat models, measured over 359 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or systemScore
151Ministral 8Blatest· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.8%
152longcat-flash-lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.8%
153MiMo V2.5 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.7%
154Llama 3.3 Nemotron Super 49B v1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.7%
155Sarvam-30B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.5%
156Granite 4.2 30B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.4%
157GPT 4.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim66.3%
158Phi-4-reasoning· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim65.8%
159Qwen3 30B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim65.8%
160DeepSeek R1 Distill Llama 70b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim65.2%
161QwQ 32B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
65.2%
162QwQ 32B Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
65.2%
163GPT-4.1 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim65%
164Gemini 2.5 Flash-Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim64.6%
165Granite 4.2 8B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim64.1%
166Nemotron Nano 9b V2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim64%
167DeepSeek R1 Distill Qwen 32B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
62.1%
168Gemini 2.0 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim62.1%
169Qwen3 Max· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim62%
170O1 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim60%
171Claude 3.5 Sonnet· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim59.4%
172DeepSeek R1 Distill Qwen 14B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
59.1%
173DeepSeek V3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim59.1%
174Gemini 1.5 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim59.1%
175Gemma 4 E4B-it· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim58.6%
176Llama 4 Scout· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim57.2%
177Phi 4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
56.1%
178Grok 2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim56%
179Granite 4.2 3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim54.8%
180Ministral 3Blatest· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim53.4%
181Aihubmix Phi 4 Mini (reasoning)· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim52%
182Qwen3.5 2B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim51.6%
183Gemini 2.0 Flash Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim51.5%
184Gemini 1.5 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim51%
185Grok 2 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim51%
186Llama 3.1 405B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim50.7%
187Llama 3.3 70B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
50.5%
188Claude 3 Opus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim50.4%
189GPT-4.1 nano· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim50.3%
190Qwen2.5 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
49.5%
191DeepSeek R1 Distill Qwen 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
49.1%
192DeepSeek R1 Distill Llama 8B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim49%
193Qwen2.5 72B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
49%
194GPT-4 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim48%
195Qwen 3 235B A22B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim47.5%
196Nova Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim46.9%
197Llama 3.1 70B Instruct0-shot· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.46.7%
198Mistral Small 3.2 24B Instruct2506· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim46.1%
199Mistral Small 3.1 24b Instruct 2503· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim46%
200Qwen 2.5 14B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim
Values from 2 sources
45.5%

Showing 151–200 of 374 results.

Not ranked

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

Model or systemScore
Qwen3.5-397B-A17B-BaseQwen3.5 release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.54.6%
Qwen3.7-Plus-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.51.5%
Qwen3.8-27B-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.45%
Qwen3.8-Flash-Next-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.51.4%

Data credits: Data by LLM Stats · 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.