StreamArena AURA Study - HR L3: leaderboard
The AURA 2.0 paper compares released models on streaming video interaction. IPIBench, StreamArena, ProactiveVideoQA and PhoStream retain their own protocols. ProactiveVideoQA reports PAUC; PhoStream’s overall score is question-weighted, and its timing categories are not guarantees of answer correctness. This board reports native evaluation accuracy (%). Higher is better.
Metric: Native evaluation accuracy (%). Source: arxiv.org. Saturation forecast: Around February 2027. 7 models tracked. Updated 14 Nov 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.5 Flash | 51.4 | #100 |
| 2 | Qwen 3.5 397B A17B | 39.3 | #251 |
| 3 | Kimi K2.6 | 38.4 | #181 |
| 4 | MiMo-V2.5 | 31.9 | #373 |
| 5 | JoyAI-VL-Interaction | 10.9 | |
| 6 | MOSS-VL-Realtime | 8.7 | |
| 7 | MiniCPM-o-4.5 | 7.6 | #1160 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=streamarena-aura-study-hr-l3 · How It Works · Data refreshed daily, snapshot 2026-11-16.