Machine Generated Text Detection on MGT benchmark Essay
100AUROCReMoDetect
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| ReMoDetectSource Model=GPT3.5 Turbo2024.05 | 100 | — | — | — | — | — | — | |
| ReMoDetectSource Model=Llama3 70B2024.05 | 100 | — | — | — | — | — | — | |
| ReMoDetectSource Model=Gemini pro2024.05 | 100 | — | — | — | — | — | — | |
| ReMoDetectSource Model=GPT4 Turbo2024.05 | 99.9 | — | — | — | — | — | — | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 99.77 | — | — | — | — | — | 95.38 | |
| ReMoDetectSource Model=Claude2024.05 | 99.7 | — | — | — | — | — | — | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 99.62 | — | — | — | — | — | 94.13 | |
| Fast-DetectGPTSource Model=GPT3.5 Turbo2024.05 | 99.6 | — | — | — | — | — | — | |
| Fast-DetectGPTSource Model=Llama3 70B2024.05 | 99.5 | — | — | — | — | — | — | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 99.39 | — | — | — | — | — | 88.31 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 99.36 | — | — | — | — | — | 95.71 | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 99.29 | — | — | — | — | — | 92.86 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 99.24 | — | — | — | — | — | 80.89 | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 99.17 | — | — | — | — | — | 87.47 | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 99.17 | — | — | — | — | — | 93.75 | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 99.13 | — | — | — | — | — | 73.6 | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 99.13 | — | — | — | — | — | 94.2 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 98.95 | — | — | — | — | — | 79.07 | |
| Fast-DetectGPTSource Model=GPT4 Turbo2024.05 | 98.9 | — | — | — | — | — | — | |
| LRRSource Model=Llama3 70B2024.05 | 98.9 | — | — | — | — | — | — | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 98.79 | — | — | — | — | — | 68.62 | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 98.65 | — | — | — | — | — | 87.61 | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 98.37 | — | — | — | — | — | 86.7 | |
| Loglik.Source Model=Llama3 70B2024.05 | 98.3 | — | — | — | — | — | — | |
| Loglik.Source Model=Gemini pro2024.05 | 98.3 | — | — | — | — | — | — | |
| Fast-DetectGPTSource Model=Gemini pro2024.05 | 98.3 | — | — | — | — | — | — | |
| LRRSource Model=GPT3.5 Turbo2024.05 | 97.8 | — | — | — | — | — | — | |
| LRRSource Model=Gemini pro2024.05 | 97.7 | — | — | — | — | — | — | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 97.61 | — | — | — | — | — | 44.71 | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 97.59 | — | — | — | — | — | 77.98 | |
| Loglik.Source Model=GPT3.5 Turbo2024.05 | 97.3 | — | — | — | — | — | — | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 97.04 | — | — | — | — | — | 75.43 | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 96.9 | — | — | — | — | — | 39.91 | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 96.8 | — | — | — | — | — | 49.82 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 96.55 | — | — | — | — | — | 62.69 | |
| Loglik.Source Model=GPT4 Turbo2024.05 | 96.5 | — | — | — | — | — | — | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 96.17 | — | — | — | — | — | 52.39 | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 96.16 | — | — | — | — | — | 46.29 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 96.03 | — | — | — | — | — | 60.44 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 95.94 | — | — | — | — | — | 13.24 | |
| RankSource Model=GPT3.5 Turbo2024.05 | 95.7 | — | — | — | — | — | — | |
| RankSource Model=Llama3 70B2024.05 | 95.3 | — | — | — | — | — | — | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 94.66 | — | — | — | — | — | 41.11 | |
| Likelihood-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 93.95 | — | — | — | — | — | 61.29 | |
| RankSource Model=GPT4 Turbo2024.05 | 93.9 | — | — | — | — | — | — | |
| LRRSource Model=GPT4 Turbo2024.05 | 93.9 | — | — | — | — | — | — | |
| RankSource Model=Gemini pro2024.05 | 93.6 | — | — | — | — | — | — | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 93.38 | — | — | — | — | — | 17.07 | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 93.08 | — | — | — | — | — | 30.67 | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 92.96 | — | — | — | — | — | 54.61 | |
| Log-Rank-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 92.96 | — | — | — | — | — | 35.91 | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 92.76 | — | — | — | — | — | 12.31 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 92.01 | — | — | — | — | — | 19.16 | |
| ChatGPT-DetectorSource Model=Llama3 70B2024.05 | 91.7 | — | — | — | — | — | — | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 91.65 | — | — | — | — | — | 14.03 | |
| Loglik.Source Model=Claude2024.05 | 91.6 | — | — | — | — | — | — | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 91.18 | — | — | — | — | — | 32.54 | |
| LikelihoodTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 90.9 | — | — | — | — | — | 20.67 | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 90.54 | — | — | — | — | — | 6.43 | |
| Log-RankTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 90.08 | — | — | — | — | — | 25.11 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 89.49 | — | — | — | — | — | 7.16 | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 89.08 | — | — | — | — | — | 20.09 | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 88.92 | — | — | — | — | — | 24.27 | |
| FastGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 88.62 | — | — | — | — | — | 16.89 | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 88.57 | — | — | — | — | — | 17.33 | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 88.43 | — | — | — | — | — | 15.07 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 86.58 | — | — | — | — | — | 3.91 | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 86.41 | — | — | — | — | — | 15.33 | |
| RankSource Model=Claude2024.05 | 85.9 | — | — | — | — | — | — | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 85.05 | — | — | — | — | — | 8.97 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 84.23 | — | — | — | — | — | 39.38 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 84.11 | — | — | — | — | — | 6.88 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 84.07 | — | — | — | — | — | 38.67 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 84.02 | — | — | — | — | — | 38.62 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 84.01 | — | — | — | — | — | 38.93 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 83.94 | — | — | — | — | — | 39.41 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 83.93 | — | — | — | — | — | 6.04 | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 83.93 | — | — | — | — | — | 39.47 | |
| OpenAI-DetectorSource Model=Llama3 70B2024.05 | 83.9 | — | — | — | — | — | — | |
| DNA-GPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 83.89 | — | — | — | — | — | 38.04 | |
| Fast-DetectGPTSource Model=Claude2024.05 | 83.6 | — | — | — | — | — | — | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 82.88 | — | — | — | — | — | 11.62 | |
| LRRSource Model=Claude2024.05 | 82.7 | — | — | — | — | — | — | |
| DetectGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT-turbo2026.02 | 82.1 | — | — | — | — | — | 3.07 | |
| ChatGPT-DetectorSource Model=GPT3.5 Turbo2024.05 | 81.5 | — | — | — | — | — | — | |
| Entropy-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 80.9 | — | — | — | — | — | 13.56 | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 80.51 | — | — | — | — | — | 17.07 | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 79.82 | — | — | — | — | — | 5.11 | |
| DetectGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 77.58 | — | — | — | — | — | 3.05 | |
| ChatGPT-DetectorSource Model=GPT4 Turbo2024.05 | 77.1 | — | — | — | — | — | — | |
| FastGPT-MTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 75.84 | — | — | — | — | — | 7.51 | |
| FastGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 75.31 | — | — | — | — | — | 3.47 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=GPT4All2026.02 | 74.19 | — | — | — | — | — | 2.73 | |
| EntropyTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Dolly2026.02 | 73.26 | — | — | — | — | — | 2.29 | |
| FastGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGLM2026.02 | 71.08 | — | — | — | — | — | 2.72 | |
| FastGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=avg2026.02 | 69.08 | — | — | — | — | — | 4.43 | |
| FastGPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=ChatGPT2026.02 | 67.68 | — | — | — | — | — | 1.64 | |
| DNA-GPTTraining Source LLM=ChatGPT-turbo, Evaluation Source LLM=Claude2026.02 | 67.53 | — | — | — | — | — | 21.64 | |
| ChatGPT-DetectorSource Model=Gemini pro2024.05 | 65.9 | — | — | — | — | — | — | |
| NPRSource Model=Gemini pro2024.05 | 65.5 | — | — | — | — | — | — |