Kwai Keye-VL-2.0 Technical Report

• URL: http://arxiv.org/abs/2606.10651v1
• Date: Tue, 09 Jun 2026 09:58:08 GMT
• Title: Kwai Keye-VL-2.0 Technical Report
• Abstract summary: Keye-VL-2.0 is a multimodal foundation model designed to advance long-video understanding and agentic intelligence.<n>It is the first to adapt DeepSeek Sparse Attention (DSA) to GQA-based multimodal architectures.<n>Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) paired with Context-RL and Video-RL overcomes the algorithmic dilemma of catastrophic forgetting.
• Score: 53.82434681649277
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce Kwai Keye-VL-2.0-30B-A3B, an open-source Mixture-of-Experts (MoE) multimodal foundation model designed to advance long-video understanding and agentic intelligence. To address the challenges of ultra-long contexts, information redundancy, and prohibitive computational costs inherent in hour-level videos, Keye-VL-2.0 is the first to adapt DeepSeek Sparse Attention (DSA) to GQA-based multimodal architectures, enabling lossless 256K context processing while capturing critical frames and long-range temporal dependencies. This architecture is underpinned by a highly optimized training and inference infrastructure, including scalable video I/O, heterogeneous ViT-LM parallelism, and custom DSA kernels that significantly maximize throughput and minimize computational overhead. Furthermore, to overcome the algorithmic dilemma of catastrophic forgetting during multi-task alignment, we introduce Cross-Modal Multi-Teacher On-Policy Distillation (MOPD) paired with Context-RL and Video-RL. By distilling dense token-level teacher feedback from on-policy rollouts back into the MoE backbone, which activates only 3B parameters, Keye-VL-2.0 natively empowers advanced agent collaboration across Code, Tool, and Search scenarios with multimodal self-correction. Extensive evaluations across video understanding, temporal grounding, reasoning, STEM, and agent benchmarks demonstrate that Keye-VL-2.0-30B-A3B achieves state-of-the-art performance among models of similar scale, particularly excelling in fine-grained temporal localization on TimeLens and long-video comprehension on Video-MME-v2 and LongVideoBench. We release our model checkpoints to accelerate community progress toward scalable and robust multimodal agentic applications.

Related papers

• Video-MME-v2: Towards the Next Stage in Benchmarks for Comprehensive Video Understanding [98.3098451637867]
  Video-MME-v2 is a comprehensive benchmark designed to rigorously evaluate the robustness and faithfulness of video understanding.<n>To guarantee data quality, Video-MME-v2 is constructed through a rigorously controlled human annotation pipeline.
  arXiv  Detail & Related papers  (2026-04-06T17:59:56Z)
• Qwen3-VL Technical Report [153.3964813640593]
  Qwen3-VL is the most capable vision-language model to date, achieving superior performance across a broad range of multimodal benchmarks.<n>It supports interleaved contexts of up to 256K tokens, seamlessly integrating text, images, and video.<n>Qwen3-VL delivers three core pillars: (i) markedly stronger pure-text understanding, surpassing comparable text-only backbones in several cases; (ii) robust long-token comprehension with a native 256K-token window for both text and interleaved multimodal inputs; and (iii) advanced multimodal reasoning across single-image, multi-image, and video tasks
  arXiv  Detail & Related papers  (2025-11-26T17:59:08Z)
• Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models [78.32948112203228]
  Video understanding represents the most challenging frontier in computer vision.<n>Recent emergence of Video-Large Multitemporal Models has demonstrated remarkable capabilities in video understanding tasks.<n>Survey aims to provide researchers and practitioners with a unified framework for advancing Video-LMM capabilities.
  arXiv  Detail & Related papers  (2025-10-06T17:10:44Z)
• SAIL-VL2 Technical Report [65.45818722427506]
  We introduce SAIL-VL2, an open-suite vision foundation model (LVM) for comprehensive multimodal understanding and reasoning.<n>SAIL-VL2 achieves state-of-the-art performance at the 2B and 8B parameter scales across diverse image and video benchmarks.
  arXiv  Detail & Related papers  (2025-09-17T14:34:02Z)
• Kwai Keye-VL 1.5 Technical Report [91.07838286692815]
  We present Keye-VL-1.5, which addresses fundamental challenges in video comprehension through three key innovations.<n>First, we introduce a novel Slow-Fast video encoding strategy that dynamically allocates computational resources based on inter-frame similarity.<n>Second, we implement a progressive four-stage pre-training methodology that systematically extends the model's context length from 8K to 128K tokens.<n>Third, we develop a comprehensive post-training pipeline focusing on reasoning enhancement and human preference alignment.
  arXiv  Detail & Related papers  (2025-09-01T15:46:58Z)
• Lumos-1: On Autoregressive Video Generation from a Unified Model Perspective [37.58855048653859]
  Autoregressive large language models (LLMs) have unified a vast range of language tasks, inspiring preliminary efforts of autoregressive video generation.<n>Lumos-1 retains autoregressive video generator architecture with minimal architectural modifications.
  arXiv  Detail & Related papers  (2025-07-11T17:59:42Z)
• VideoLights: Feature Refinement and Cross-Task Alignment Transformer for Joint Video Highlight Detection and Moment Retrieval [8.908777234657046]
  Large-language and vision-language models (LLM/LVLMs) have gained prominence across various domains.<n>Here we propose VideoLights, a novel HD/MR framework addressing these limitations through (i) Convolutional Projection and Feature Refinement modules.<n> Comprehensive experiments on QVHighlights, TVSum, and Charades-STA benchmarks demonstrate state-of-the-art performance.
  arXiv  Detail & Related papers  (2024-12-02T14:45:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.