Skip to content

Benchmark

CritPT

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

Models measured
156
Scores recorded
268
Sources
3
IQ weight
2.67

Score distribution

0.98%
Mean
0.106%
Median
4%
Std dev
0%
10th percentile
0.314%
90th percentile
0%
Min
0%31.4%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.67 for chat models, measured over 156 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201MiniMax M2.7· Epoch AI Benchmarking Hub data0.571%
202Qwen3.5 35B A3Bnone· Epoch AI Benchmarking Hub data0.571%
203Qwen3.5 9Bnone· Epoch AI Benchmarking Hub data0.571%
204Claude Opus 4· Epoch AI Benchmarking Hub data0.3%
205Claude Sonnet 4· Epoch AI Benchmarking Hub data0.286%
206Command A Plus 05 2026· Epoch AI Benchmarking Hub data0.286%
207DeepSeek V4.1 Flashnone· Epoch AI Benchmarking Hub data0.286%
208DeepSeek V4.1 Flash 1017none· Epoch AI Benchmarking Hub data0.286%
209GPT-5.6 Lunanone· Epoch AI Benchmarking Hub data0.286%
210Magistral Medium· Epoch AI Benchmarking Hub data0.286%
211Magistral Small 2509· Epoch AI Benchmarking Hub data0.286%
212o3 Minihigh· Epoch AI Benchmarking Hub data0.286%
213Qwen 3.8 27Bnone· Epoch AI Benchmarking Hub data0.286%
214Qwen3 30B A3B2507, thinking· Epoch AI Benchmarking Hub data0.286%
215Qwen3 32B· Epoch AI Benchmarking Hub data0.286%
216Qwen3.5 9B· Epoch AI Benchmarking Hub data0.286%
217Qwen3.6 35B A3B· Epoch AI Benchmarking Hub data0.286%
218C4AI Command A 03 2025· Epoch AI Benchmarking Hub data0%
219Claude Haiku 3.5· Epoch AI Benchmarking Hub data0%
220Claude Haiku 4.5· Epoch AI Benchmarking Hub data0%
221Cogito v2.1 671B· Epoch AI Benchmarking Hub data0%
222DeepSeek V3· Epoch AI Benchmarking Hub data0%
223DeepSeek V4 Pro 0813none· Epoch AI Benchmarking Hub data0%
224Devstral Small 2505· Epoch AI Benchmarking Hub data0%
225Gemini 3.5 Flash Lite· Epoch AI Benchmarking Hub data0%
226Gemma 3 12B IT· Epoch AI Benchmarking Hub data0%
227Gemma 3 270M It· Epoch AI Benchmarking Hub data0%
228Gemma 3 27B IT· Epoch AI Benchmarking Hub data0%
229Gemma 4 26B A4B· Epoch AI Benchmarking Hub data0%
230GPT 4o· Epoch AI Benchmarking Hub data0%
231GPT 5minimal· Epoch AI Benchmarking Hub data0%
232GPT OSS 120Blow· Epoch AI Benchmarking Hub data0%
233GPT OSS 20blow· Epoch AI Benchmarking Hub data0%
234GPT-4.1 mini· Epoch AI Benchmarking Hub data0%
235GPT-4.1 nano· Epoch AI Benchmarking Hub data0%
236GPT-5 mini· Epoch AI Benchmarking Hub data0%
237GPT-5.5 Instant· Epoch AI Benchmarking Hub data0%
238Grok 4.3· Epoch AI Benchmarking Hub data0%
239Grok 4.3none· Epoch AI Benchmarking Hub data0%
240Llama 3.1 405B Instruct· Epoch AI Benchmarking Hub data0%
241Llama 3.1 8B Instruct· Epoch AI Benchmarking Hub data0%
242Llama 3.2 11B Vision Instruct· Epoch AI Benchmarking Hub data0%
243Llama 3.3 70B Instruct· Epoch AI Benchmarking Hub data0%
244Llama 4 Maverick· Epoch AI Benchmarking Hub data0%
245Llama 4 Scout· Epoch AI Benchmarking Hub data0%
246Mercury 2.5· Epoch AI Benchmarking Hub data0%
247MiMo V2 Flash· Epoch AI Benchmarking Hub data0%
248Mistral Large 3· Epoch AI Benchmarking Hub data0%
249Mistral Medium2604· Epoch AI Benchmarking Hub data0%
250Mistral Medium 3.1· Epoch AI Benchmarking Hub data0%

Showing 201–250 of 268 results.

Data credits: Data by LLM Stats · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · 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.