EnviroExam
Models answer multiple choice questions from 42 environmental science courses. The composite index discounts mean accuracy when performance varies across courses.
The EnvGPT team evaluated five models on EnviroExam. The original release has 210 development questions and 726 test questions. Environmental biotechnology accounts for 20 questions across the two splits.
- Made by
- Harbin Institute of Technology
- Released
- May 2024
- Kind
- Science questions
- Level
- University
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Language
- English
- Interaction
- Single turn
- Items from
- Exams and contests
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Biology and medicine
- Strongest model tested
-
Gemma 7B #211 on the capability index
- Last model added
- February 2024
- Scores refreshed
- November 20th, 2026
deepseek-67b-chat tops the leaderboard with 84.9 Qwen1.5-7B-Chat leads with 80.6 among models released by February 2024
Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. Where a model ran with several settings or harnesses, we keep its best. Every run is in the next chart. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Qwen1.5-7B-Chat, at 80.6. This view shows only models we track, each at its best.
The newest model with a score here came out in February 2024, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 84.9 deepseek-67b-chat
- 84.5 llama-3-70b-instruct
- 80.8 mixtral-8x7b-instruct-v0.1
- 80.6
Qwen1.5-7B-Chat
- 79.8 internlm2-chat-7b
- 78.6 llama-3-8b-instruct
- 78.2 internlm2-chat-20b
- 75.6 internlm-chat-20b
- 75.6
Qwen1.5-14B-Chat
- 75.1
Yi-6B-Chat
- 74.7
Qwen-14B
- 73.2 Baichuan2-13B-Chat
- 71.3 chatglm3-6b-32k
- 69.8 qwen-7b-chat
- 68.1 bluelm-7b-chat
- 67.4 ChatGLM3-6B
- 67.3 internlm-chat-7b
- 67.1 baichuan-13b-chat
- 66.7 llama-2-70b-chat
- 63.1 deepseek-7b-chat
- 59.4
Gemma 7B
- 59.2 tigerbot-13b-chat-v2
- 59.1 vicuna-13b-v1.5
- 53.3 vicuna-7b-v1.5-16k
- 51.2 tigerbot-7b-chat-v3
- 49.1
Mistral-7B-Instruct-v0.1
- 47.8 llama-2-13b-chat
- 47.3 gemma-2b-it
- 46.0 chatglm2-6b
- 17.8 skywork-13b
- 15.4 vicuna-7b-v1.5
- 1 deepseek-67b-chat Five-shot 84.9
- 2 llama-3-70b-instruct Five-shot 84.5
- 3 mixtral-8x7b-instruct-v0.1 Five-shot 80.8
- 4
Qwen1.5-7B-Chat Five-shot 80.6
- 5 internlm2-chat-7b Five-shot 79.8
- 6 llama-3-8b-instruct Five-shot 78.6
- 7 internlm2-chat-20b Five-shot 78.2
- 8 internlm-chat-20b Five-shot 75.6
- 9
Qwen1.5-14B-Chat Five-shot 75.6
- 10
Yi-6B-Chat Five-shot 75.1
- 11
Qwen-14B Five-shot 74.7
- 12 Baichuan2-13B-Chat Zero-shot 73.2
- 13 chatglm3-6b-32k Five-shot 71.3
- 14 qwen-7b-chat Five-shot 69.8
- 15 bluelm-7b-chat Zero-shot 68.1
- 16 ChatGLM3-6B Five-shot 67.4
- 17 internlm-chat-7b Zero-shot 67.3
- 18 baichuan-13b-chat Zero-shot 67.1
- 19 llama-2-70b-chat Five-shot 66.7
- 20 deepseek-7b-chat Five-shot 63.1
- 21
Gemma 7B Zero-shot 59.4
- 22 tigerbot-13b-chat-v2 Zero-shot 59.2
- 23 vicuna-13b-v1.5 Five-shot 59.1
- 24 vicuna-7b-v1.5-16k Five-shot 53.3
- 25 tigerbot-7b-chat-v3 Zero-shot 51.2
- 26
Mistral-7B-Instruct-v0.1 Five-shot 49.1
- 27 llama-2-13b-chat Five-shot 47.8
- 28 gemma-2b-it Zero-shot 47.3
- 29 chatglm2-6b Zero-shot 46.0
- 30 skywork-13b Zero-shot 17.8
- 31 vicuna-7b-v1.5 Zero-shot 15.4
Source: The EnviroExam paper
- 1
Qwen1.5-7B-Chat Five-shot 80.6
- 2 internlm-chat-20b Five-shot 75.6
- 3
Qwen1.5-14B-Chat Five-shot 75.6
- 4
Yi-6B-Chat Five-shot 75.1
- 5
Qwen-14B Five-shot 74.7
- 6 Baichuan2-13B-Chat Zero-shot 73.2
- 7 ChatGLM3-6B Five-shot 67.4
- 8
Gemma 7B Zero-shot 59.4
- 9
Mistral-7B-Instruct-v0.1 Five-shot 49.1
- 10 chatglm2-6b Zero-shot 46.0
Source: The EnviroExam paper
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | deepseek-67b-chat | 84.9 | Five-shot | Official leaderboard | |
| 2 | llama-3-70b-instruct | 84.5 | Five-shot | Official leaderboard | |
| 3 | mixtral-8x7b-instruct-v0.1 | 80.8 | Five-shot | Official leaderboard | |
| 4 | | 80.6 | Five-shot | Feb 2024 | Official leaderboard |
| 5 | internlm2-chat-7b | 79.8 | Five-shot | Official leaderboard | |
| 6 | llama-3-8b-instruct | 78.6 | Five-shot | Official leaderboard | |
| 7 | internlm2-chat-20b | 78.2 | Five-shot | Official leaderboard | |
| 8 | internlm-chat-20b | 75.6 | Five-shot | Sep 2023 | Official leaderboard |
| 9 | | 75.6 | Five-shot | Feb 2024 | Official leaderboard |
| 10 | | 75.1 | Five-shot | Nov 2023 | Official leaderboard |
| 11 | | 74.7 | Five-shot | Sep 2023 | Official leaderboard |
| 12 | Baichuan2-13B-Chat | 73.2 | Zero-shot | Sep 2023 | Official leaderboard |
| 13 | chatglm3-6b-32k | 71.3 | Five-shot | Official leaderboard | |
| 14 | qwen-7b-chat | 69.8 | Five-shot | Official leaderboard | |
| 15 | bluelm-7b-chat | 68.1 | Zero-shot | Official leaderboard | |
| 16 | ChatGLM3-6B | 67.4 | Five-shot | Oct 2023 | Official leaderboard |
| 17 | internlm-chat-7b | 67.3 | Zero-shot | Official leaderboard | |
| 18 | baichuan-13b-chat | 67.1 | Zero-shot | Official leaderboard | |
| 19 | llama-2-70b-chat | 66.7 | Five-shot | Official leaderboard | |
| 20 | deepseek-7b-chat | 63.1 | Five-shot | Official leaderboard | |
| 21 | | 59.4 | Zero-shot | Feb 2024 | Official leaderboard |
| 22 | tigerbot-13b-chat-v2 | 59.2 | Zero-shot | Official leaderboard | |
| 23 | vicuna-13b-v1.5 | 59.1 | Five-shot | Official leaderboard | |
| 24 | vicuna-7b-v1.5-16k | 53.3 | Five-shot | Official leaderboard | |
| 25 | tigerbot-7b-chat-v3 | 51.2 | Zero-shot | Official leaderboard | |
| 26 | | 49.1 | Five-shot | Sep 2023 | Official leaderboard |
| 27 | llama-2-13b-chat | 47.8 | Five-shot | Official leaderboard | |
| 28 | gemma-2b-it | 47.3 | Zero-shot | Official leaderboard | |
| 29 | chatglm2-6b | 46.0 | Zero-shot | Jun 2023 | Official leaderboard |
| 30 | skywork-13b | 17.8 | Zero-shot | Official leaderboard | |
| 31 | vicuna-7b-v1.5 | 15.4 | Zero-shot | Official leaderboard |
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | | 80.6 | Five-shot | Feb 2024 | Official leaderboard |
| 2 | internlm-chat-20b | 75.6 | Five-shot | Sep 2023 | Official leaderboard |
| 3 | | 75.6 | Five-shot | Feb 2024 | Official leaderboard |
| 4 | | 75.1 | Five-shot | Nov 2023 | Official leaderboard |
| 5 | | 74.7 | Five-shot | Sep 2023 | Official leaderboard |
| 6 | Baichuan2-13B-Chat | 73.2 | Zero-shot | Sep 2023 | Official leaderboard |
| 7 | ChatGLM3-6B | 67.4 | Five-shot | Oct 2023 | Official leaderboard |
| 8 | | 59.4 | Zero-shot | Feb 2024 | Official leaderboard |
| 9 | | 49.1 | Five-shot | Sep 2023 | Official leaderboard |
| 10 | chatglm2-6b | 46.0 | Zero-shot | Jun 2023 | Official leaderboard |
Rankings by part
We read 43 parts from the benchmark’s paper, each ranking models on one part of the test, such as environmental science, environmental fluid mechanics and environmental analytical chemistry. Baichuan2-13B-Chat leads 9 of them, and 8 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.
environmental science
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 100%
- 2
Gemma 7B ↑6 100%
- 3 Baichuan2-13B-Chat ↑3 94.4%
- 4 internlm-chat-20b ↓2 94.4%
- 5
Qwen-14B 94.4%
- 6
Qwen1.5-7B-Chat ↓5 94.4%
- 7
Yi-6B-Chat ↓3 88.9%
- 8
Mistral-7B-Instruct-v0.1 ↑1 83.3%
- 9 ChatGLM3-6B ↓2 77.8%
- 10 chatglm2-6b 55.6%
environmental fluid mechanics
10 models we track have a score here.
- 1 chatglm2-6b ↑9 100%
- 2 Baichuan2-13B-Chat ↑4 100%
- 3
Qwen-14B ↑2 100%
- 4 internlm-chat-20b ↓2 83.3%
- 5
Yi-6B-Chat ↓1 83.3%
- 6
Qwen1.5-14B-Chat ↓3 83.3%
- 7
Qwen1.5-7B-Chat ↓6 83.3%
- 8 ChatGLM3-6B ↓1 66.7%
- 9
Mistral-7B-Instruct-v0.1 50%
- 10
Gemma 7B ↓2 50%
environmental analytical chemistry
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 100%
- 2 internlm-chat-20b 83.3%
- 3
Qwen1.5-14B-Chat 83.3%
- 4
Qwen1.5-7B-Chat ↓3 83.3%
- 5
Qwen-14B 66.7%
- 6 ChatGLM3-6B ↑1 66.7%
- 7
Yi-6B-Chat ↓3 66.7%
- 8
Gemma 7B 66.7%
- 9 chatglm2-6b ↑1 50%
- 10
Mistral-7B-Instruct-v0.1 ↓1 16.7%
principles and applications of ecology
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 100%
- 2 internlm-chat-20b 100%
- 3
Yi-6B-Chat ↑1 100%
- 4
Qwen-14B ↑1 92.9%
- 5
Qwen1.5-7B-Chat ↓4 92.9%
- 6
Gemma 7B ↑2 92.9%
- 7
Mistral-7B-Instruct-v0.1 ↑2 85.7%
- 8 ChatGLM3-6B ↓1 85.7%
- 9
Qwen1.5-14B-Chat ↓6 85.7%
- 10 chatglm2-6b 64.3%
pollution control microbiology
10 models we track have a score here.
- 1
Qwen1.5-7B-Chat 100%
- 2
Qwen-14B ↑3 88.9%
- 3
Yi-6B-Chat ↑1 88.9%
- 4 Baichuan2-13B-Chat ↑2 83.3%
- 5 ChatGLM3-6B ↑2 83.3%
- 6
Qwen1.5-14B-Chat ↓3 83.3%
- 7
Gemma 7B ↑1 72.2%
- 8 internlm-chat-20b ↓6 66.7%
- 9
Mistral-7B-Instruct-v0.1 50%
- 10 chatglm2-6b 44.4%
atmospheric pollution control engineering
10 models we track have a score here.
- 1
Qwen-14B ↑4 95.2%
- 2 internlm-chat-20b 90.5%
- 3
Qwen1.5-14B-Chat 90.5%
- 4 Baichuan2-13B-Chat ↑2 85.7%
- 5
Yi-6B-Chat ↓1 81%
- 6
Qwen1.5-7B-Chat ↓5 81%
- 7
Gemma 7B ↑1 81%
- 8 ChatGLM3-6B ↓1 71.4%
- 9
Mistral-7B-Instruct-v0.1 61.9%
- 10 chatglm2-6b 52.4%
environmental monitoring
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 100%
- 2 internlm-chat-20b 100%
- 3
Yi-6B-Chat ↑1 100%
- 4
Qwen-14B ↑1 95.5%
- 5 ChatGLM3-6B ↑2 95.5%
- 6
Qwen1.5-14B-Chat ↓3 95.5%
- 7
Qwen1.5-7B-Chat ↓6 95.5%
- 8
Gemma 7B 86.4%
- 9 chatglm2-6b ↑1 77.3%
- 10
Mistral-7B-Instruct-v0.1 ↓1 68.2%
environmental impact assessment
10 models we track have a score here.
- 1
Yi-6B-Chat ↑3 87.5%
- 2
Qwen1.5-14B-Chat ↑1 87.5%
- 3 Baichuan2-13B-Chat ↑3 83.3%
- 4
Qwen1.5-7B-Chat ↓3 83.3%
- 5
Qwen-14B 75%
- 6
Mistral-7B-Instruct-v0.1 ↑3 75%
- 7 ChatGLM3-6B 75%
- 8 internlm-chat-20b ↓6 70.8%
- 9
Gemma 7B ↓1 58.3%
- 10 chatglm2-6b 41.7%
urban water supply and drainage system engineering
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Qwen-14B ↑3 100%
- 3
Qwen1.5-7B-Chat ↓2 100%
- 4 Baichuan2-13B-Chat ↑2 88.9%
- 5 ChatGLM3-6B ↑2 88.9%
- 6
Gemma 7B ↑2 88.9%
- 7
Yi-6B-Chat ↓3 77.8%
- 8
Qwen1.5-14B-Chat ↓5 77.8%
- 9 chatglm2-6b ↑1 66.7%
- 10
Mistral-7B-Instruct-v0.1 ↓1 55.6%
environmental soil science
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Qwen-14B ↑3 100%
- 3 ChatGLM3-6B ↑4 100%
- 4
Qwen1.5-14B-Chat ↓1 100%
- 5
Qwen1.5-7B-Chat ↓4 100%
- 6 Baichuan2-13B-Chat 77.8%
- 7
Mistral-7B-Instruct-v0.1 ↑2 77.8%
- 8
Yi-6B-Chat ↓4 77.8%
- 9
Gemma 7B ↓1 77.8%
- 10 chatglm2-6b 66.7%
environmental economics
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 94.7%
- 2 internlm-chat-20b 84.2%
- 3
Qwen-14B ↑2 84.2%
- 4
Yi-6B-Chat 84.2%
- 5
Qwen1.5-7B-Chat ↓4 84.2%
- 6 Baichuan2-13B-Chat 79%
- 7 ChatGLM3-6B 73.7%
- 8
Gemma 7B 73.7%
- 9
Mistral-7B-Instruct-v0.1 68.4%
- 10 chatglm2-6b 52.6%
environmental decision support system
10 models we track have a score here.
- 1
Qwen-14B ↑4 100%
- 2
Mistral-7B-Instruct-v0.1 ↑7 100%
- 3
Yi-6B-Chat ↑1 100%
- 4
Qwen1.5-14B-Chat ↓1 100%
- 5
Qwen1.5-7B-Chat ↓4 100%
- 6 Baichuan2-13B-Chat 94.4%
- 7 internlm-chat-20b ↓5 94.4%
- 8 ChatGLM3-6B ↓1 94.4%
- 9
Gemma 7B ↓1 88.9%
- 10 chatglm2-6b 83.3%
solid waste treatment and resource utilization
10 models we track have a score here.
- 1
Qwen-14B ↑4 94.4%
- 2
Yi-6B-Chat ↑2 94.4%
- 3 Baichuan2-13B-Chat ↑3 88.9%
- 4
Qwen1.5-14B-Chat ↓1 88.9%
- 5
Qwen1.5-7B-Chat ↓4 88.9%
- 6
Gemma 7B ↑2 88.9%
- 7 internlm-chat-20b ↓5 83.3%
- 8
Mistral-7B-Instruct-v0.1 ↑1 83.3%
- 9 ChatGLM3-6B ↓2 72.2%
- 10 chatglm2-6b 50%
water pollution control engineering
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Yi-6B-Chat ↑2 100%
- 3
Qwen-14B ↑2 87.5%
- 4 ChatGLM3-6B ↑3 87.5%
- 5
Qwen1.5-14B-Chat ↓2 87.5%
- 6 Baichuan2-13B-Chat 75%
- 7
Mistral-7B-Instruct-v0.1 ↑2 75%
- 8
Qwen1.5-7B-Chat ↓7 75%
- 9 chatglm2-6b ↑1 62.5%
- 10
Gemma 7B ↓2 50%
environmental planning and management
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 90%
- 2 internlm-chat-20b 90%
- 3
Qwen-14B ↑2 90%
- 4
Qwen1.5-7B-Chat ↓3 90%
- 5 ChatGLM3-6B ↑2 85%
- 6
Yi-6B-Chat ↓2 85%
- 7
Qwen1.5-14B-Chat ↓4 85%
- 8 chatglm2-6b ↑2 70%
- 9
Gemma 7B ↓1 65%
- 10
Mistral-7B-Instruct-v0.1 ↓1 60%
indoor environmental pollution and control
10 models we track have a score here.
- 1
Yi-6B-Chat ↑3 100%
- 2 Baichuan2-13B-Chat ↑4 95.7%
- 3
Qwen-14B ↑2 95.7%
- 4
Qwen1.5-14B-Chat ↓1 95.7%
- 5
Qwen1.5-7B-Chat ↓4 95.7%
- 6 internlm-chat-20b ↓4 91.3%
- 7
Mistral-7B-Instruct-v0.1 ↑2 87%
- 8 ChatGLM3-6B ↓1 87%
- 9 chatglm2-6b ↑1 73.9%
- 10
Gemma 7B ↓2 73.9%
environmental statistics
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 94.7%
- 2
Mistral-7B-Instruct-v0.1 ↑7 89.5%
- 3
Qwen1.5-14B-Chat 89.5%
- 4 internlm-chat-20b ↓2 84.2%
- 5 ChatGLM3-6B ↑2 84.2%
- 6
Yi-6B-Chat ↓2 84.2%
- 7
Qwen1.5-7B-Chat ↓6 84.2%
- 8
Qwen-14B ↓3 79%
- 9
Gemma 7B ↓1 73.7%
- 10 chatglm2-6b 68.4%
industrial ecology
10 models we track have a score here.
- 1 ChatGLM3-6B ↑6 95.5%
- 2 internlm-chat-20b 90.9%
- 3
Qwen-14B ↑2 90.9%
- 4
Yi-6B-Chat 90.9%
- 5
Qwen1.5-14B-Chat ↓2 90.9%
- 6
Qwen1.5-7B-Chat ↓5 90.9%
- 7 Baichuan2-13B-Chat ↓1 86.4%
- 8
Mistral-7B-Instruct-v0.1 ↑1 81.8%
- 9
Gemma 7B ↓1 81.8%
- 10 chatglm2-6b 40.9%
environmental physical pollution and control
10 models we track have a score here.
- 1
Qwen-14B ↑4 95%
- 2
Qwen1.5-14B-Chat ↑1 95%
- 3
Qwen1.5-7B-Chat ↓2 95%
- 4
Yi-6B-Chat 90%
- 5 internlm-chat-20b ↓3 85%
- 6 ChatGLM3-6B ↑1 85%
- 7
Gemma 7B ↑1 85%
- 8 chatglm2-6b ↑2 80%
- 9 Baichuan2-13B-Chat ↓3 80%
- 10
Mistral-7B-Instruct-v0.1 ↓1 75%
environmental nanotechnology
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 100%
- 2 ChatGLM3-6B ↑5 100%
- 3
Yi-6B-Chat ↑1 100%
- 4
Qwen1.5-7B-Chat ↓3 100%
- 5 internlm-chat-20b ↓3 92.9%
- 6
Qwen-14B ↓1 92.9%
- 7
Mistral-7B-Instruct-v0.1 ↑2 92.9%
- 8
Qwen1.5-14B-Chat ↓5 92.9%
- 9 chatglm2-6b ↑1 71.4%
- 10
Gemma 7B ↓2 71.4%
specialty wastewater treatment
10 models we track have a score here.
- 1
Qwen1.5-7B-Chat 77.8%
- 2
Gemma 7B ↑6 77.8%
- 3 internlm-chat-20b ↓1 66.7%
- 4 ChatGLM3-6B ↑3 66.7%
- 5
Yi-6B-Chat ↓1 66.7%
- 6 Baichuan2-13B-Chat 55.6%
- 7
Qwen-14B ↓2 55.6%
- 8
Qwen1.5-14B-Chat ↓5 55.6%
- 9 chatglm2-6b ↑1 44.4%
- 10
Mistral-7B-Instruct-v0.1 ↓1 33.3%
modern biology
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 94.4%
- 2 Baichuan2-13B-Chat ↑4 88.9%
- 3
Yi-6B-Chat ↑1 88.9%
- 4
Qwen1.5-7B-Chat ↓3 88.9%
- 5 internlm-chat-20b ↓3 83.3%
- 6
Qwen-14B ↓1 83.3%
- 7 ChatGLM3-6B 77.8%
- 8 chatglm2-6b ↑2 66.7%
- 9
Mistral-7B-Instruct-v0.1 66.7%
- 10
Gemma 7B ↓2 66.7%
theory and technology of environmental pollution prevention and control
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Qwen-14B ↑3 100%
- 3
Yi-6B-Chat ↑1 100%
- 4
Qwen1.5-14B-Chat ↓1 100%
- 5 Baichuan2-13B-Chat ↑1 93.8%
- 6
Mistral-7B-Instruct-v0.1 ↑3 93.8%
- 7
Qwen1.5-7B-Chat ↓6 93.8%
- 8 chatglm2-6b ↑2 81.3%
- 9 ChatGLM3-6B ↓2 81.3%
- 10
Gemma 7B ↓2 75%
reaction kinetics and reactor design
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Qwen-14B ↑3 90%
- 3 ChatGLM3-6B ↑4 90%
- 4
Yi-6B-Chat 90%
- 5
Qwen1.5-14B-Chat ↓2 90%
- 6 Baichuan2-13B-Chat 80%
- 7
Mistral-7B-Instruct-v0.1 ↑2 80%
- 8
Qwen1.5-7B-Chat ↓7 80%
- 9
Gemma 7B ↓1 70%
- 10 chatglm2-6b 10%
engineering ethics and environmental ethics
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 100%
- 2
Qwen1.5-14B-Chat ↑1 95.7%
- 3 internlm-chat-20b ↓1 91.3%
- 4
Qwen-14B ↑1 91.3%
- 5
Yi-6B-Chat ↓1 91.3%
- 6
Mistral-7B-Instruct-v0.1 ↑3 87%
- 7 ChatGLM3-6B 87%
- 8
Qwen1.5-7B-Chat ↓7 87%
- 9
Gemma 7B ↓1 82.6%
- 10 chatglm2-6b 69.6%
sludge treatment disposal and resource utilization
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 92.9%
- 2
Mistral-7B-Instruct-v0.1 ↑7 92.9%
- 3 internlm-chat-20b ↓1 85.7%
- 4
Qwen-14B ↑1 85.7%
- 5
Yi-6B-Chat ↓1 85.7%
- 6
Qwen1.5-7B-Chat ↓5 85.7%
- 7 ChatGLM3-6B 78.6%
- 8
Qwen1.5-14B-Chat ↓5 78.6%
- 9
Gemma 7B ↓1 71.4%
- 10 chatglm2-6b 50%
safe disposal and resource recovery of hazardous waste
10 models we track have a score here.
- 1
Qwen-14B ↑4 100%
- 2 Baichuan2-13B-Chat ↑4 94.4%
- 3 internlm-chat-20b ↓1 94.4%
- 4
Mistral-7B-Instruct-v0.1 ↑5 94.4%
- 5
Yi-6B-Chat ↓1 94.4%
- 6
Qwen1.5-14B-Chat ↓3 94.4%
- 7
Qwen1.5-7B-Chat ↓6 94.4%
- 8
Gemma 7B 83.3%
- 9 chatglm2-6b ↑1 77.8%
- 10 ChatGLM3-6B ↓3 77.8%
theory and technology of air pollution prevention and control
10 models we track have a score here.
- 1
Yi-6B-Chat ↑3 100%
- 2 ChatGLM3-6B ↑5 94.1%
- 3 Baichuan2-13B-Chat ↑3 88.2%
- 4 internlm-chat-20b ↓2 88.2%
- 5
Qwen1.5-14B-Chat ↓2 88.2%
- 6
Qwen1.5-7B-Chat ↓5 88.2%
- 7
Gemma 7B ↑1 88.2%
- 8
Qwen-14B ↓3 82.3%
- 9 chatglm2-6b ↑1 76.5%
- 10
Mistral-7B-Instruct-v0.1 ↓1 76.5%
environmental geographic information system
10 models we track have a score here.
- 1
Mistral-7B-Instruct-v0.1 ↑8 95.8%
- 2 Baichuan2-13B-Chat ↑4 91.7%
- 3 internlm-chat-20b ↓1 91.7%
- 4
Qwen-14B ↑1 91.7%
- 5
Qwen1.5-14B-Chat ↓2 91.7%
- 6
Qwen1.5-7B-Chat ↓5 91.7%
- 7
Yi-6B-Chat ↓3 87.5%
- 8 ChatGLM3-6B ↓1 79.2%
- 9 chatglm2-6b ↑1 66.7%
- 10
Gemma 7B ↓2 62.5%
clean production and energy conservation and emissions reduction
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2
Qwen1.5-14B-Chat ↑1 100%
- 3
Qwen-14B ↑2 94.1%
- 4
Yi-6B-Chat 94.1%
- 5
Qwen1.5-7B-Chat ↓4 94.1%
- 6 Baichuan2-13B-Chat 82.3%
- 7 ChatGLM3-6B 82.3%
- 8
Gemma 7B 82.3%
- 9
Mistral-7B-Instruct-v0.1 64.7%
- 10 chatglm2-6b 52.9%
circulate economy and industrial ecology methodologies
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 95.5%
- 2 Baichuan2-13B-Chat ↑4 90.9%
- 3 internlm-chat-20b ↓1 90.9%
- 4
Qwen-14B ↑1 90.9%
- 5
Qwen1.5-7B-Chat ↓4 90.9%
- 6
Yi-6B-Chat ↓2 86.4%
- 7
Gemma 7B ↑1 86.4%
- 8 chatglm2-6b ↑2 81.8%
- 9 ChatGLM3-6B ↓2 81.8%
- 10
Mistral-7B-Instruct-v0.1 ↓1 77.3%
atmospheric particulate matter detection and analysis techniques
10 models we track have a score here.
- 1
Yi-6B-Chat ↑3 100%
- 2
Qwen1.5-14B-Chat ↑1 100%
- 3 Baichuan2-13B-Chat ↑3 95%
- 4
Qwen-14B ↑1 95%
- 5
Qwen1.5-7B-Chat ↓4 95%
- 6
Gemma 7B ↑2 95%
- 7 internlm-chat-20b ↓5 85%
- 8 ChatGLM3-6B ↓1 80%
- 9
Mistral-7B-Instruct-v0.1 70%
- 10 chatglm2-6b 65%
environmental electrochemical theory and technology
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 76.9%
- 2
Qwen-14B ↑3 76.9%
- 3 Baichuan2-13B-Chat ↑3 69.2%
- 4 ChatGLM3-6B ↑3 69.2%
- 5
Qwen1.5-14B-Chat ↓2 69.2%
- 6
Qwen1.5-7B-Chat ↓5 69.2%
- 7 chatglm2-6b ↑3 61.5%
- 8
Mistral-7B-Instruct-v0.1 ↑1 61.5%
- 9
Yi-6B-Chat ↓5 61.5%
- 10
Gemma 7B ↓2 61.5%
environmental microwave chemistry technology
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 85.7%
- 2 Baichuan2-13B-Chat ↑4 78.6%
- 3
Qwen1.5-7B-Chat ↓2 78.6%
- 4 internlm-chat-20b ↓2 71.4%
- 5
Qwen-14B 71.4%
- 6 ChatGLM3-6B ↑1 71.4%
- 7
Yi-6B-Chat ↓3 71.4%
- 8
Gemma 7B 64.3%
- 9 chatglm2-6b ↑1 50%
- 10
Mistral-7B-Instruct-v0.1 ↓1 42.9%
advanced environmental chemistry
10 models we track have a score here.
- 1 internlm-chat-20b ↑1 100%
- 2 Baichuan2-13B-Chat ↑4 95%
- 3
Qwen-14B ↑2 95%
- 4
Yi-6B-Chat 95%
- 5
Qwen1.5-14B-Chat ↓2 95%
- 6
Qwen1.5-7B-Chat ↓5 95%
- 7
Gemma 7B ↑1 95%
- 8 ChatGLM3-6B ↓1 90%
- 9 chatglm2-6b ↑1 75%
- 10
Mistral-7B-Instruct-v0.1 ↓1 60%
mass transfer processes
10 models we track have a score here.
- 1 ChatGLM3-6B ↑6 75%
- 2
Qwen1.5-7B-Chat ↓1 75%
- 3 Baichuan2-13B-Chat ↑3 68.8%
- 4
Qwen-14B ↑1 68.8%
- 5
Yi-6B-Chat ↓1 68.8%
- 6
Gemma 7B ↑2 68.8%
- 7 internlm-chat-20b ↓5 56.3%
- 8
Qwen1.5-14B-Chat ↓5 56.3%
- 9
Mistral-7B-Instruct-v0.1 50%
- 10 chatglm2-6b 37.5%
carbon neutrality theory and technology
10 models we track have a score here.
- 1
Qwen-14B ↑4 95.5%
- 2
Yi-6B-Chat ↑2 95.5%
- 3
Qwen1.5-14B-Chat 95.5%
- 4 Baichuan2-13B-Chat ↑2 90.9%
- 5 internlm-chat-20b ↓3 90.9%
- 6 ChatGLM3-6B ↑1 90.9%
- 7
Qwen1.5-7B-Chat ↓6 90.9%
- 8
Mistral-7B-Instruct-v0.1 ↑1 81.8%
- 9 chatglm2-6b ↑1 72.7%
- 10
Gemma 7B ↓2 72.7%
environmental biotechnology
10 models we track have a score here.
- 1
Qwen1.5-7B-Chat 73.3%
- 2
Qwen-14B ↑3 66.7%
- 3
Yi-6B-Chat ↑1 66.7%
- 4
Qwen1.5-14B-Chat ↓1 66.7%
- 5 chatglm2-6b ↑5 60%
- 6 Baichuan2-13B-Chat 60%
- 7 internlm-chat-20b ↓5 60%
- 8
Mistral-7B-Instruct-v0.1 ↑1 60%
- 9 ChatGLM3-6B ↓2 60%
- 10
Gemma 7B ↓2 53.3%
modern detection techniques
10 models we track have a score here.
- 1
Qwen1.5-14B-Chat ↑2 96.5%
- 2
Qwen1.5-7B-Chat ↓1 96.5%
- 3 ChatGLM3-6B ↑4 93.1%
- 4
Yi-6B-Chat 93.1%
- 5
Gemma 7B ↑3 93.1%
- 6 internlm-chat-20b ↓4 89.7%
- 7
Qwen-14B ↓2 89.7%
- 8 Baichuan2-13B-Chat ↓2 86.2%
- 9 chatglm2-6b ↑1 69%
- 10
Mistral-7B-Instruct-v0.1 ↓1 51.7%
environmental pollution investigation and tracing
10 models we track have a score here.
- 1 Baichuan2-13B-Chat ↑5 95.7%
- 2
Qwen-14B ↑3 95.7%
- 3 chatglm2-6b ↑7 87%
- 4 internlm-chat-20b ↓2 87%
- 5
Mistral-7B-Instruct-v0.1 ↑4 87%
- 6
Qwen1.5-7B-Chat ↓5 87%
- 7
Yi-6B-Chat ↓3 82.6%
- 8
Qwen1.5-14B-Chat ↓5 82.6%
- 9 ChatGLM3-6B ↓2 73.9%
- 10
Gemma 7B ↓2 39.1%
persistent organic pollutants
10 models we track have a score here.
- 1
Qwen-14B ↑4 100%
- 2
Qwen1.5-14B-Chat ↑1 95.2%
- 3 Baichuan2-13B-Chat ↑3 90.5%
- 4
Qwen1.5-7B-Chat ↓3 90.5%
- 5 internlm-chat-20b ↓3 85.7%
- 6
Yi-6B-Chat ↓2 85.7%
- 7 ChatGLM3-6B 81%
- 8
Gemma 7B 81%
- 9
Mistral-7B-Instruct-v0.1 61.9%
- 10 chatglm2-6b 33.3%
environmental behavior of atmospheric science and climate change
10 models we track have a score here.
- 1
Qwen-14B ↑4 100%
- 2
Yi-6B-Chat ↑2 100%
- 3 Baichuan2-13B-Chat ↑3 94.1%
- 4 internlm-chat-20b ↓2 94.1%
- 5
Qwen1.5-14B-Chat ↓2 94.1%
- 6
Qwen1.5-7B-Chat ↓5 94.1%
- 7
Mistral-7B-Instruct-v0.1 ↑2 82.3%
- 8 ChatGLM3-6B ↓1 82.3%
- 9
Gemma 7B ↓1 82.3%
- 10 chatglm2-6b 70.6%
M*(1-cv)
10 models we track have a score here.
- 1
Qwen1.5-7B-Chat 80.6%
- 2 internlm-chat-20b 75.6%
- 3
Qwen1.5-14B-Chat 75.6%
- 4
Yi-6B-Chat 75.1%
- 5
Qwen-14B 74.7%
- 6 Baichuan2-13B-Chat 73.2%
- 7 ChatGLM3-6B 67.4%
- 8
Gemma 7B 59.4%
- 9
Mistral-7B-Instruct-v0.1 49.1%
- 10 chatglm2-6b 46%
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The EnviroExam paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Five-shot | 80.6 | ||||
| internlm-chat-20b | Five-shot | 75.6 | |||
| Five-shot | 75.6 | ||||
| Five-shot | 75.1 | ||||
| Five-shot | 74.7 | ||||
| Baichuan2-13B-Chat | Zero-shot | 73.2 | |||
| Zero-shot | 71.0 | ||||
| Zero-shot | 70.7 | ||||
| internlm-chat-20b | Zero-shot | 69.4 | |||
| ChatGLM3-6B | Five-shot | 67.4 | |||
| Baichuan2-13B-Chat | Five-shot | 66.6 | |||
| Zero-shot | 65.3 | ||||
| Zero-shot | 65.1 | ||||
| ChatGLM3-6B | Zero-shot | 62.7 | |||
| Zero-shot | 59.4 | ||||
| Five-shot | 49.1 | ||||
| Zero-shot | 47.2 | ||||
| chatglm2-6b | Zero-shot | 46.0 | |||
| Five-shot | 31.5 | ||||
| chatglm2-6b | Five-shot | 3.1 |
Scores at each effort level
Models the leaderboard ran at two or more effort levels. We don’t have enough cost data here to say what the extra effort costs.
Source: The EnviroExam paper
| Model | Effort | Harness | Score | Cost per task | Output tokens |
|---|---|---|---|---|---|
| Five-shot | 80.6 | ||||
| internlm-chat-20b | Five-shot | 75.6 | |||
| Five-shot | 75.6 | ||||
| Five-shot | 75.1 | ||||
| Five-shot | 74.7 | ||||
| Baichuan2-13B-Chat | Zero-shot | 73.2 | |||
| Zero-shot | 71.0 | ||||
| Zero-shot | 70.7 | ||||
| internlm-chat-20b | Zero-shot | 69.4 | |||
| ChatGLM3-6B | Five-shot | 67.4 | |||
| Baichuan2-13B-Chat | Five-shot | 66.6 | |||
| Zero-shot | 65.3 | ||||
| Zero-shot | 65.1 | ||||
| ChatGLM3-6B | Zero-shot | 62.7 | |||
| Zero-shot | 59.4 | ||||
| Five-shot | 49.1 | ||||
| Zero-shot | 47.2 | ||||
| chatglm2-6b | Zero-shot | 46.0 | |||
| Five-shot | 31.5 | ||||
| chatglm2-6b | Five-shot | 3.1 |
Number one over time
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: The EnviroExam paper
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official EnviroExam leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Science and reasoning
- GPQA Diamond In use 198 graduate-level 'Google-proof' multiple-choice questions in biology, chemistry and physics.
Gemini 4 Carbon 97%
- Terminal-Bench-Science In use 70 scientific research-workflow tasks across five science areas, run in Harbor containers. Hidden tests give the resolution rate. GPT-6 Astra 68.1%
- BioMysteryBench In use Bioinformatics problems on anonymized real datasets that require real analysis with tools. Rubric-graded, with human-solvable and human-difficult splits. Claude Fable 5.2 83.7%
- FrontierScience In use Models solve physics, chemistry, and biology problems in an Olympiad track graded by short answer and a Research track graded by 10-point rubrics.
Intern-S2-397B 87%
- HiPhO In use Models solve theory problems from 2024 and 2025 physics olympiad exams, some with diagrams. Official marking schemes grade answers and steps, mapped to medal levels.
Gemini 3 Pro 78.3%
- LABBench2 In use 11 biology research sub-tasks (literature, figures, tables, patents, trials and more) with tool access. Macro-averaged accuracy. GPT-5.2 60.9%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 2 years ago, in February 2024, for Gemma 7B. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the benchmark’s paper (31 entries).
- Pass Models scored 10 models we track, plus 21 other entries on its leaderboard.
- Warning Newest score The newest came 2 years ago, in February 2024, for Gemma 7B.
- Note Frontier models None of the top 10 has a score. The strongest tested is Gemma 7B (#211).
- Note Headroom Scored as an index, with no fixed ceiling.
Leaderboard import 3 pass
- Pass Importer We read the board on huggingface.co: 20 rows for models we track and 21 other entries at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.