# Long-term dialogue memory on LongMemEval (test) Source: [Long-term dialogue memory on LongMemEval (test)]() Collection index: [llms.txt]() Data updated: 2026-03-05T07:21:08.477Z Evaluation of the model's long-term dialogue memory capabilities on the test split of the LongMemEval benchmark, measured by accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Long-term dialogue memory | | Dataset (including any stated split) | LongMemEval (test) | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. - [Information Density](): Higher is better; unit: SCALAR. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | MemBuilder | 85.75 | Not recorded | 69647fe753fdd5c361744743 | | MemBuilder | 85.5 | Not recorded | 69647fe753fdd5c361744743 | | MemBuilder | 84.25 | Not recorded | 69647fe753fdd5c361744743 | | MemBuilder | 78.5 | Not recorded | 69647fe753fdd5c361744743 | | MemBuilder | 76 | Not recorded | 69647fe753fdd5c361744743 | | PLUGMEM | 75.1 | Not recorded | 69a90ddd2d603e6de3e4058a | | MIRIX | 73.25 | Not recorded | 69647fe753fdd5c361744743 | | LiCoMemory | 73 | Not recorded | 69a90ddd2d603e6de3e4058a | | Zep | 71.2 | Not recorded | 69a90ddd2d603e6de3e4058a | | RAG-Utterance | 69 | Not recorded | 69647fe753fdd5c361744743 | | RAG-Session | 66.75 | Not recorded | 69647fe753fdd5c361744743 | | Vanilla Retrieval | 63.6 | Not recorded | 69a90ddd2d603e6de3e4058a | | MemBuilder | 62.75 | Not recorded | 69647fe753fdd5c361744743 | | All Context | 62.4 | Not recorded | 69a90ddd2d603e6de3e4058a | | A-Mem | 61 | Not recorded | 69a90ddd2d603e6de3e4058a | | MemBuilder | 56 | Not recorded | 69647fe753fdd5c361744743 | | Mem0 | 47 | Not recorded | 69647fe753fdd5c361744743 | | No Context | 14.8 | Not recorded | 69a90ddd2d603e6de3e4058a | ### MemBuilder Accuracy: 85.75. Result ID: 6996957d5698665ec4dedd97. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"Qwen3-4B","Training stage":"SFT + RL","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 85.5. Result ID: 6996957d5698665ec4dedd93. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"Claude 4.5 Sonnet","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 84.25. Result ID: 6996957d5698665ec4dedd95. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"Qwen3-4B","Training stage":"SFT","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 78.5. Result ID: 6996957d5698665ec4dedd91. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"GPT-4.1","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 76. Result ID: 6996957d5698665ec4dedd92. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"QwQ-32B","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### PLUGMEM Accuracy: 75.1. Result ID: 69a92ed76dbb7075b95227a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Emb":"NVE","LLM":"Q32 + Q72","#TokAvg.":"362.58"}. This record does not assert independent reproduction. ### MIRIX Accuracy: 73.25. Result ID: 6996957d5698665ec4dedd90. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction method":"Prompting-based","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### LiCoMemory Accuracy: 73. Result ID: 69a92ed76dbb7075b95227a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Emb":"NVE","LLM":"4o + Q72","#TokAvg.":"5914.85"}. This record does not assert independent reproduction. ### Zep Accuracy: 71.2. Result ID: 69a92ed76dbb7075b95227a6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Emb":"BGE-m3","LLM":"4o","#TokAvg.":"1600"}. This record does not assert independent reproduction. ### RAG-Utterance Accuracy: 69. Result ID: 6996957d5698665ec4dedd8e. [Reporting paper]() Recorded evaluation configuration: {"Memory construction method":"Retrieval-based","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### RAG-Session Accuracy: 66.75. Result ID: 6996957d5698665ec4dedd8d. [Reporting paper]() Recorded evaluation configuration: {"Memory construction method":"Retrieval-based","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### Vanilla Retrieval Accuracy: 63.6. Result ID: 69a92ed76dbb7075b95227a4. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Emb":"NVE","LLM":"Q72","#TokAvg.":"3742.52"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 62.75. Result ID: 6996957d5698665ec4dedd96. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"Qwen3-4B","Training stage":"RL","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### All Context Accuracy: 62.4. Result ID: 69a92ed76dbb7075b95227a3. [Reporting paper]() Recorded evaluation configuration: {"Emb":"-","LLM":"Q72","#TokAvg.":"107K"}. This record does not assert independent reproduction. ### A-Mem Accuracy: 61. Result ID: 69a92ed76dbb7075b95227a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Emb":"NVE","LLM":"4o + Q72","#TokAvg.":"4225.85"}. This record does not assert independent reproduction. ### MemBuilder Accuracy: 56. Result ID: 6996957d5698665ec4dedd94. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction model":"Qwen3-4B","Training stage":"Base","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### Mem0 Accuracy: 47. Result ID: 6996957d5698665ec4dedd8f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Memory construction method":"Prompting-based","Answer model":"Claude 4.5 Sonnet"}. This record does not assert independent reproduction. ### No Context Accuracy: 14.8. Result ID: 69a92ed76dbb7075b95227a2. [Reporting paper]() Recorded evaluation configuration: {"Emb":"-","LLM":"Q72","#TokAvg.":"-"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.