Memory-augmented Question Answering on LongMemEval
92.9SSU ScoreZep
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| ZepLLM Model=GPT-4o-mini2026.05 | 92.9 | 47.4 | 53.3 | 54.1 | 74.4 | 75 | 63.21 | 3.8 | 5.45 | |
| ZepLLM Model=GPT-4o2026.05 | 92.9 | 57.9 | 56.7 | 62.4 | 83.3 | 80.4 | 70.4 | 3.8 | 5.45 | |
| VikingMemLLM Model=GPT-4o2026.05 | 92.9 | 69.92 | 56.7 | 66.92 | 76.92 | 98.21 | 75.8 | 0.25 | 0.89 | |
| VikingMemLLM Model=GPT-4o-mini2026.05 | 90 | 53.31 | 46.67 | 55.89 | 71.23 | 96.43 | 66.36 | 0.25 | 0.89 | |
| Mem0-graphLLM Model=GPT-4o2026.05 | 84.29 | 48.87 | 30 | 41.35 | 75.64 | 48.21 | 54.8 | 1.1 | 2.09 | |
| Full-ContextLLM Model=GPT-4o-mini2026.05 | 81.43 | 40.6 | 30 | 36.84 | 76.92 | 82.14 | 55 | — | — | |
| Full-ContextLLM Model=GPT-4o2026.05 | 81.43 | 44.36 | 20 | 45.11 | 78.21 | 94.64 | 59.2 | — | — | |
| Mem0LLM Model=GPT-4o2026.05 | 80 | 50.38 | 26.67 | 42.11 | 74.36 | 10.71 | 50.2 | 0.4 | 1 | |
| Mem0-graphLLM Model=GPT-4o-mini2026.05 | 72.86 | 38.35 | 16.67 | 34.59 | 60.26 | 35.71 | 44 | 1.1 | 2.09 | |
| Mem0LLM Model=GPT-4o-mini2026.05 | 68.57 | 41.35 | 20 | 37.59 | 66.67 | 5.36 | 42.8 | 0.4 | 1 |