# TTF: Temporal Token Fusion for Efficient Video-Language Model

arXiv: https://arxiv.org/abs/2605.07355 · HTML: https://mlsanity.com/paper/2605.07355/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 11 of its 21 references that are in the map.
- First posted on arXiv 8 May 2026.

## Builds on (most cited first, 11 of 11)

- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis](https://mlsanity.com/paper/2405.21075/) (2024)
- [LongVideoBench: A Benchmark for Long-context Interleaved Video-Language Understanding](https://mlsanity.com/paper/2407.15754/) (2024)
- [LMMs-Eval: Reality Check on the Evaluation of Large Multimodal Models](https://mlsanity.com/paper/2407.12772/) (2024)
- [VisionZip: Longer is Better but Not Necessary in Vision Language Models](https://mlsanity.com/paper/2412.04467/) (2024)
- [MLVU: Benchmarking Multi-task Long Video Understanding](https://mlsanity.com/paper/2406.04264/) (2024)
- [DyCoke: Dynamic Compression of Tokens for Fast Video Large Language Models](https://mlsanity.com/paper/2411.15024/) (2024)
- [Challenges and Responses in the Practice of Large Language Models](https://mlsanity.com/paper/2408.09416/) (2024)
- [Efficient and Robust Long-Form Speech Recognition with Hybrid H3-Conformer](https://mlsanity.com/paper/2410.04159/) (2024)
- [Context-AI Tunes: Context-Aware AI-Generated Music for Stress Reduction](https://mlsanity.com/paper/2505.09872/) (2025)
- [MaMe & MaRe: Matrix-Based Token Merging and Restoration for Efficient Visual Perception and Synthesis](https://mlsanity.com/paper/2604.13432/) (2026)

## Abstract (arXiv)

Video-language models (VLMs) face rapid inference costs as visual token counts scale with video length. For example, 32 frames at 448×448 resolution already yield >8,000 visual tokens in Qwen3-VL, making LLM prefill the dominant throughput bottleneck. Existing methods often rely on global similarity or attention-guided compression, incurring offsets to their gains. We propose Temporal Token Fusion (TTF), a training-free, plug-and-play pre-LLM token compression framework that exploits structured temporal redundancy in video. TTF automatically selects an anchor frame, then for each subsequent frame, performs a local window similarity search (e.g.,3× 3), fusing tokens that exceed a threshold. The compressed sequence maintains positional consistency across both prefill and decoding through coordinate realignment, enabling seamless integration with existing VLM pipelines. On Qwen3-VL-8B with threshold t=0.70, TTF removes about 67% of visual tokens while retaining 99.5% of the baseline accuracy and introducing only ≈0.16,GFLOPs of matching overhead. Overall, TTF offers a practical, efficient solution for video understanding. The code is available at hrefhttps://github.com/Cominder/ttfhttps://github.com/Cominder/ttf