| OneStreamer: Unifying Perception, Memory, and Proactive Response in Streaming Video Interaction | | 86.9 | | OneStreamer | 2026-10-01 |
| Supervising the Visible State: Separating Representation from View Selection in Streaming Video | ✓ Link | 86.33 | | ViSS + TVS | 2026-10-26 |
| Memory as Weights: Internalizing Long-Term History for Streaming Videos | ✓ Link | 84.8 | | MaLoW (Qwen3-VL-4B) | 2026-10-15 |
| Supervising the Visible State: Separating Representation from View Selection in Streaming Video | ✓ Link | 84.77 | | ViSS (recent-4) | 2026-10-26 |
| StreamOPD: A Post-Training Recipe with Spatio-Temporal Cue Gating for Streaming Video Understanding | ✓ Link | 84.55 | | ST-CueGate (Qwen3.5-4B, n=4) (extra data) | 2026-08-17 |
| MEMO: Multi-Level Entity-Aware Memory for Streaming Video Understanding | ✓ Link | 83.7 | | Qwen3-VL-8B + MEMO | 2026-09-30 |
| What Should a Streaming Video Model Remember? | | 82.67 | | SelectStream (Qwen3-VL-8B) | 2026-06-15 |
| Memory as Weights: Internalizing Long-Term History for Streaming Videos | ✓ Link | 82.3 | | MaLoW (Qwen3.5-9B) | 2026-10-15 |
| What Should a Streaming Video Model Remember? | | 81.42 | | SelectStream (Qwen2.5-VL-7B) | 2026-06-15 |
| StreamTTT: Reconciling Real-Time Perception and Long-Term Memory in Streaming VLMs | ✓ Link | 81.32 | | StreamTTT-4B | 2026-08-13 |
| Memory as Weights: Internalizing Long-Term History for Streaming Videos | ✓ Link | 81.2 | | MaLoW (VideoChat3-4B) | 2026-10-15 |
| Memory as Weights: Internalizing Long-Term History for Streaming Videos | ✓ Link | 81.0 | | MaLoW (MOSS-VL-Real-11B) | 2026-10-15 |
| StreamPro: From Reactive Perception to Proactive Decision-Making in Streaming Video | | 79.3 | | StreamPro-SFT (4B) | 2026-05-11 |
| Foresight: planning future perception in streaming VLMs without retraining | | 79.30 | 66.00 | Foresight | 2026-10-02 |
| StreamPro: From Reactive Perception to Proactive Decision-Making in Streaming Video | | 78.9 | | StreamPro-GRPO (4B) | 2026-05-11 |
| MEMO: Multi-Level Entity-Aware Memory for Streaming Video Understanding | ✓ Link | 78.5 | | Qwen2.5-VL-7B + MEMO | 2026-09-30 |
| OASIS: On-Demand Hierarchical Event Memory for Streaming Video Reasoning | | 78.22 | | OASIS (Qwen3-VL-8B) | 2026-04-18 |
| StreamForest: Efficient Online Video Understanding with Persistent Event Memory | | 77.3 | | StreamForest-7B | 2025-09-29 |
| StreamPro: From Reactive Perception to Proactive Decision-Making in Streaming Video | | 77.1 | | StreamPro-SFT (3B) | 2026-05-11 |
| GROVE: Growing and Reasoning over Temporally Stratified Memory from Streaming Video Experience | | 77.0 | | GROVE | 2026-08-03 |
| StreamPro: From Reactive Perception to Proactive Decision-Making in Streaming Video | | 76.0 | | StreamPro-GRPO (3B) | 2026-05-11 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 75.69 | 67.07 | Gemini 1.5 Pro | 2024-11-06 |
| TimeChat-Online: 80% Visual Tokens are Naturally Redundant in Streaming Videos | | 75.28 | | TimeChat-Online-7B | 2025-04-24 |
| Thinking in Streaming Video | | 75.00 | | ThinkStream-3B | 2026-03-13 |
| DeltaStreamer: Towards Human-Like Agentic Streaming Video Understanding | | 74.80 | | DeltaStreamer | 2026-10-14 |
| Thinking in Streaming Video | | 73.31 | | Qwen2.5-VL-7B | 2026-03-13 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 73.28 | 60.15 | GPT-4o | 2024-11-06 |
| MEMO: Multi-Level Entity-Aware Memory for Streaming Video Understanding | ✓ Link | 73.2 | | LLaVA-OneVision-7B + MEMO | 2026-09-30 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 72.44 | 57.68 | Claude 3.5 Sonnet | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 71.12 | 56.36 | LLaVA-OneVision-7B | 2024-11-06 |
| StreamKV: Streaming Video Question-Answering with Segment-based KV Cache Retrieval and Compression | | 71.0 | 58.9 | StreamKV-7B | 2025-11-10 |
| OASIS: On-Demand Hierarchical Event Memory for Streaming Video Reasoning | | 70.59 | | OASIS (Qwen2.5-VL-7B) | 2026-04-18 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 69.04 | 54.14 | Qwen2-VL-7B | 2024-11-06 |
| Dispider: Enabling Video LLMs with Active Real-Time Interaction via Disentangled Perception, Decision, and Reaction | | 67.63 | 53.12 | Dispider-7B | 2025-01-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 67.44 | 53.85 | MiniCPM-V 2.6 | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 66.96 | 52.77 | LLaVA-NeXT-Video-32B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 64.60 | 51.10 | Kangaroo-7B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 63.72 | 51.40 | InternVL2-8B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 59.96 | 48.66 | LongVA-7B | 2024-11-06 |
| MEMO: Multi-Level Entity-Aware Memory for Streaming Video Understanding | ✓ Link | 59.5 | | LLaVA-OneVision-0.5B + MEMO | 2026-09-30 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 53.96 | 42.53 | Video-CCAM-14B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 52.32 | 43.20 | VILA-1.5-8B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 49.52 | 40.40 | Video-LLaMA2-7B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 35.99 | 32.48 | VideoLLM-online-8B | 2024-11-06 |
| StreamingBench: Assessing the Gap for MLLMs to Achieve Streaming Video Understanding | | 23.23 | 24.04 | Flash-VStream-7B | 2024-11-06 |