PhoStream AURA Study - Partly Correct: 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 forward timing rate (%). Higher is better.

Metric: Native forward timing rate (%). Source: arxiv.org. Saturation forecast: Around June 2027. 13 models tracked. Updated 14 Nov 2026.

Top models

#ModelScoreOverall rank
1Seed-1.656.85#453
2Seed 1.841.18#245
3MOSS-VL-Realtime29.62
4Gemini 3 Pro20.77#136
5Qwen 3.5 9B15.77#814
6Qwen2.5-Omni-7B13.85#1866
7MMDuet212.24
8Qwen 3 VL 8B11.06#899
9Dispider10.28#4338
10Qwen 3 VL 30B A3B7.49#794
11JoyAI-VL-Interaction3.75
12Qwen 3 Omni 30B A3B2.03#870
13VideoLLM-online-8B0.04

No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.

Interactive version: theaggregate.ai/benchmark?slug=phostream-aura-study-partly-correct · How It Works · Data refreshed daily, snapshot 2026-11-16.