StreamProfileBench: leaderboard
Large Language Models (LLMs) have reshaped user profiling, yet current evaluations mainly focus on static data snapshots. This paradigm overlooks the reality of personalized systems, where User-Generated Content (UGC) arrives continuously and fine-grained profile evolve rapidly. To bridge this gap, we introduce StreamProfileBench, a large-scale benchmark for fine-grained streaming user profiling.
Metric: F1_NS Overall (self-reported). Source: benchmarklist.com. Saturation forecast: Around April 2027. 14 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash | 54.97 | #99 |
| 2 | DeepSeek V3.2 | 44.63 | #205 |
| 3 | GLM-4.7 | 44.21 | #192 |
| 4 | GPT-5.1 | 42.73 | #135 |
| 5 | Qwen 3 14B | 41.33 | #537 |
| 6 | Qwen 3 32B | 40.82 | #443 |
| 7 | Llama 3.1 70B | 40.38 | #592 |
| 8 | MiniMax-M2.5 | 37.9 | #303 |
| 9 | GPT-OSS-120B | 37.12 | #341 |
| 10 | GPT-5 Mini | 35.44 | #186 |
| 11 | GPT-OSS-20B | 34.68 | #518 |
| 12 | GPT-4o Mini | 33.53 | #599 |
| 13 | Qwen 3 8B | 31.54 | #674 |
| 14 | Llama 3.1 8B | 18.85 | #1151 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=streamprofilebench · How It Works · Data refreshed daily, snapshot 2026-10-20.