# FrameThinker: Learning to Think with Long Videos via Multi-Turn Frame Spotlighting

arXiv: https://arxiv.org/abs/2509.24304 · HTML: https://mlsanity.com/paper/2509.24304/

## Key facts

- Cited by 42 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 27 of its 45 references that are in the map.
- First posted on arXiv 29 Sep 2025.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Video understanding with MLLMs](https://mlsanity.com/mllm/video-understanding-with-mllms/)

## Builds on (most cited first, 27 of 27)

- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Gemini 2.5: Pushing the Frontier with Advanced Reasoning, Multimodality, Long Context, and Next Generation Agentic Capabilities](https://mlsanity.com/paper/2507.06261/) (2025)
- [DAPO: An Open-Source LLM Reinforcement Learning System at Scale](https://mlsanity.com/paper/2503.14476/) (2025)
- [Gemini: A Family of Highly Capable Multimodal Models](https://mlsanity.com/paper/2312.11805/) (2023)
- [SFT Memorizes, RL Generalizes: A Comparative Study of Foundation Model Post-training](https://mlsanity.com/paper/2501.17161/) (2025)
- [Long Context Transfer from Language to Vision](https://mlsanity.com/paper/2406.16852/) (2024)
- [Video-R1: Reinforcing Video Reasoning in MLLMs](https://mlsanity.com/paper/2503.21776/) (2025)
- [LVBench: An Extreme Long Video Understanding Benchmark](https://mlsanity.com/paper/2406.08035/) (2024)
- [DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning](https://mlsanity.com/paper/2505.14362/) (2025)
- [PLLaVA : Parameter-free LLaVA Extension from Images to Videos for Video Dense Captioning](https://mlsanity.com/paper/2404.16994/) (2024)
- [LongVILA: Scaling Long-Context Visual Language Models for Long Videos](https://mlsanity.com/paper/2408.10188/) (2024)
- [Learning to Reason under Off-Policy Guidance](https://mlsanity.com/paper/2504.14945/) (2025)
- [Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learning](https://mlsanity.com/paper/2505.15966/) (2025)
- [VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning](https://mlsanity.com/paper/2504.06958/) (2025)
- [STAR: A Benchmark for Situated Reasoning in Real-World Videos](https://mlsanity.com/paper/2405.09711/) (2024)
- [Thinking with Images for Multimodal Reasoning: Foundations, Methods, and Future Frontiers](https://mlsanity.com/paper/2506.23918/) (2025)
- [VITA-1.5: Towards GPT-4o Level Real-Time Vision and Speech Interaction](https://mlsanity.com/paper/2501.01957/) (2025)
- [A Survey of Reinforcement Learning for Large Reasoning Models](https://mlsanity.com/paper/2509.08827/) (2025)
- [OpenThinkIMG: Learning to Think with Images via Visual Tool Reinforcement Learning](https://mlsanity.com/paper/2505.08617/) (2025)
- [A Survey of Efficient Reasoning for Large Reasoning Models: Language, Multimodality, and Beyond](https://mlsanity.com/paper/2503.21614/) (2025)
- [Video-Holmes: Can MLLM Think Like Holmes for Complex Video Reasoning?](https://mlsanity.com/paper/2505.21374/) (2025)
- [Deep Video Discovery: Agentic Search with Tool Use for Long-form Video Understanding](https://mlsanity.com/paper/2505.18079/) (2025)
- [Scaling RL to Long Videos](https://mlsanity.com/paper/2507.07966/) (2025)
- [GRPO-CARE: Consistency-Aware Reinforcement Learning for Multimodal Reasoning](https://mlsanity.com/paper/2506.16141/) (2025)
- [VideoMind: A Chain-of-LoRA Agent for Temporal-Grounded Video Reasoning](https://mlsanity.com/paper/2503.13444/) (2025)
- [Does Your Vision-Language Model Get Lost in the Long Video Sampling Dilemma?](https://mlsanity.com/paper/2503.12496/) (2025)

## Built on by (most cited first, 30 of 42)

- [Spotlight on Token Perception for Multimodal Reinforcement Learning](https://mlsanity.com/paper/2510.09285/) (2025)
- [Active Video Perception: Iterative Evidence Seeking for Agentic Long Video Understanding](https://mlsanity.com/paper/2512.05774/) (2025)
- [Skyra: AI-Generated Video Detection via Grounded Artifact Reasoning](https://mlsanity.com/paper/2512.15693/) (2025)
- [GEMS: Agent-Native Multimodal Generation with Memory and Skills](https://mlsanity.com/paper/2603.28088/) (2026)
- [DiffThinker: Towards Generative Multimodal Reasoning with Diffusion Models](https://mlsanity.com/paper/2512.24165/) (2025)
- [Conan: Progressive Learning to Reason Like a Detective over Multi-Scale Visual Evidence](https://mlsanity.com/paper/2510.20470/) (2025)
- [Video-LMM Post-Training: A Deep Dive into Video Reasoning with Large Multimodal Models](https://mlsanity.com/paper/2510.05034/) (2025)
- [GIFT: Global Irreplaceability Frame Targeting for Efficient Video Understanding](https://mlsanity.com/paper/2603.25072/) (2026)
- [ResAdapt: Adaptive Resolution for Efficient Multimodal Reasoning](https://mlsanity.com/paper/2603.28610/) (2026)
- [VideoBrain: Learning Adaptive Frame Sampling for Long Video Understanding](https://mlsanity.com/paper/2602.04094/) (2026)
- [VideoTemp-o3: Harmonizing Temporal Grounding and Video Understanding in Agentic Thinking-with-Videos](https://mlsanity.com/paper/2602.07801/) (2026)
- [VideoZeroBench: Probing the Limits of Video MLLMs with Spatio-Temporal Evidence Verification](https://mlsanity.com/paper/2604.01569/) (2026)
- [EVA: Efficient Reinforcement Learning for End-to-End Video Agent](https://mlsanity.com/paper/2603.22918/) (2026)
- [VideoSEAL: Mitigating Evidence Misalignment in Agentic Long Video Understanding by Decoupling Answer Authority](https://mlsanity.com/paper/2605.12571/) (2026)
- [Agent Planning Benchmark: A Diagnostic Framework for Planning Capabilities in LLM Agents](https://mlsanity.com/paper/2606.04874/) (2026)
- [DeepSport: A Multimodal Large Language Model for Comprehensive Sports Video Reasoning via Agentic Reinforcement Learning](https://mlsanity.com/paper/2511.12908/) (2025)
- [PRPO: Perception-Reinforced Policy Optimization via Token-Level Dynamic Advantage Reshaping](https://mlsanity.com/paper/2606.08708/) (2026)
- [VideoSearcher: Empowering Video Deep Research with Multi-Tool Agentic Reasoning via Reinforcement Learning](https://mlsanity.com/paper/2607.02927/) (2026)
- [Persistent Visual Memory: Sustaining Perception for Deep Generation in LVLMs](https://mlsanity.com/paper/2605.00814/) (2026)
- [Watch, Remember, Reason: Human-View Video Understanding with MLLMs](https://mlsanity.com/paper/2606.07433/) (2026)
- [When Thinking Hurts: Mitigating Visual Forgetting in Video Reasoning via Frame Repetition](https://mlsanity.com/paper/2603.16256/) (2026)
- [Query-Conditioned Evidential Keyframe Sampling for MLLM-Based Long-Form Video Understanding](https://mlsanity.com/paper/2604.01002/) (2026)
- [Find, Fix, Reason: Context Repair for Video Reasoning](https://mlsanity.com/paper/2604.16243/) (2026)
- [OmniReasoner: Thinking with Long Audio-Video via Native Tool Use](https://mlsanity.com/paper/2607.19339/) (2026)
- [Thinking with Drafts: Speculative Temporal Reasoning for Efficient Long Video Understanding](https://mlsanity.com/paper/2512.00805/) (2025)
- [MedScope: Incentivizing "Think with Videos" for Clinical Reasoning via Coarse-to-Fine Tool Calling](https://mlsanity.com/paper/2602.13332/) (2026)
- [Towards Sparse Video Understanding and Reasoning](https://mlsanity.com/paper/2602.13602/) (2026)
- [Bridging Visual Representation and Reinforcement Learning from Verifiable Rewards in Large Vision-Language Models](https://mlsanity.com/paper/2603.27375/) (2026)
- [Incentivizing Vision Language Models to Search for Long Video Question Answering](https://mlsanity.com/paper/2607.02959/) (2026)
- [MedClaw: Heuristic Agent Harness for Long-Horizon Surgical Video Reasoning](https://mlsanity.com/paper/2608.14015/) (2026)

## Abstract (arXiv)

While Large Vision-Language Models (LVLMs) have achieved substantial progress in video understanding, their application to long video reasoning is hindered by uniform frame sampling and static textual reasoning, which are inefficient and struggle to handle visually intensive video tasks. To overcome these challenges, in this paper, we introduce the concept of thinking with long videos and propose a novel framework FrameThinker. Within this framework, LVLMs are able to iteratively interrogate video content. Developing such video reasoning capabilities in LVLMs presents notable challenges, particularly in adapting the model to new video actions (e.g. select frame), and designing reward functions to guide LVLMs to adopt the newly introduced action. To solve these challenges, we propose a two-phase training strategy, first employing Supervised Fine-Tuning (SFT) to instill fundamental action capabilities, followed by Reinforcement Learning (RL) to optimize a strategic decision-making policy. Notably, in this RL phase, we conduct an in-depth and comprehensive exploration of the reward design for each action and format reward. Extensive experiments on reasoning benchmarks like Video-Holmes, LongVideo-Reason, and long-video understanding benchmarks such as LongVideoBench, MLVU, VideoMME, and LVBench, demonstrate that FrameThinker achieves a significant average improvement of +10.4% over baselines while drastically reducing the number of processed frames. Most notably, our 7B model, FrameThinker establishes a new state-of-the-art on LongVideo-Reason, achieving 76.1% accuracy using an average of only 20.6 frames. This not only outperforms the competitive LongVILA-R1 (72.0%) but does so with over 20x fewer frames (vs. 512), demonstrating unparalleled efficiency and effectiveness.