Skip to content
Papers.

Video models are zero-shot learners and reasoners research paper by Google DeepMind, 2025

Google DeepMind · Sep 24, 2025 · Multimodal and robotics · 215 citations · 100 upvotes · unverified

Read on arXiv

What it shows

Veo 3, a generative video model, exhibits zero-shot capabilities across various visual tasks, suggesting a trajectory towards becoming a unified, generalist vision foundation model.

UnverifiedHugging Face's summary; not yet checked by hand.

More from Google DeepMind

All 12
PaperCitations
DiffusionGemma Technical ReportDiffusionGemma is a fine-tuned mixture-of-experts language model that uses discrete diffusion to generate text blocks in parallel, achieving high speed while preserving capabilities like multimodal inputs and reasoning.Foundation models · Jul 2026 · Unverified1
Gemma 4 Technical ReportOpen multimodal models from 2.3B to 31B parameters, with a thinking mode and image and audio input.Foundation models · Jul 2026110
LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMsLiteFrame, a lightweight video encoder with Compressed Token Distillation training method, reduces latency and increases frame processing capacity for long-form video understanding in Video LLMs while maintaining accuracy.Inference and efficiency · May 2026 · Unverified1
Understanding the Challenges in Iterative Generative Optimization with LLMsGenerative optimization using large language models faces challenges due to implicit design decisions about artifact modification and learning evidence that significantly impact success across different applications.Foundation models · Mar 2026 · Unverified8
LoGeR: Long-Context Geometric Reconstruction with Hybrid MemoryLoGeR enables long-term 3D video reconstruction by combining bidirectional priors with a hybrid memory system that includes parametric Test-Time Training and non-parametric sliding window attention mechanisms.Reasoning · Mar 2026 · Unverified35
SIMA 2: A Generalist Embodied Agent for Virtual WorldsSIMA 2, built on a Gemini foundation model, interacts in 3D virtual worlds, reasons about goals, handles complex instructions, and autonomously learns new skills through open-ended self-improvement.Agents and evaluation · Dec 2025 · Unverified18
Robot Learning from a Physical World ModelPhysWorld integrates video generation and physical world modeling to enable accurate robotic manipulation from visual demonstrations without real robot data.Multimodal and robotics · Nov 2025 · Unverified21
Vibe Checker: Aligning Code Evaluation with Human PreferenceVibe Checker evaluates LLMs by combining functional correctness and instruction following to better align with human coding preferences.Agents and evaluation · Oct 2025 · Unverified2

Other multimodal and robotics papers

Topic
PaperCitations
Segment AnythingA promptable model and a dataset of over a billion masks that cut out any object in any image.Meta · Apr 202315.7k
GPT-4o System CardGPT-4o is an omnimodal autoregressive model trained to handle text, audio, image, and video inputs, offering high-performance outputs across these modalities, with particular strengths in vision and audio.OpenAI · Oct 2024 · Unverified4,980
SAM 3: Segment Anything with ConceptsSegment Anything Model 3 achieves state-of-the-art performance in promptable concept segmentation and tracking by leveraging a unified model architecture with decoupled recognition and localization.Meta · Nov 2025 · Unverified999
DeepSeek-VL: Towards Real-World Vision-Language UnderstandingDeepSeek-VL is an open-source vision-language model that achieves state-of-the-art performance in real-world applications by combining a hybrid vision encoder with effective pretraining strategies to preserve language model capabilities.DeepSeek · Mar 2024 · Unverified889
Janus: Decoupling Visual Encoding for Unified Multimodal Understanding and GenerationJanus, an autoregressive framework with separate visual encoding pathways within a unified transformer architecture, enhances performance in unified multimodal understanding and generation.DeepSeek · Oct 2024 · Unverified481
SAM 3D: 3Dfy Anything in ImagesSAM 3D is a generative model that reconstructs 3D objects from single images using a multi-stage training framework that includes synthetic pretraining and real-world alignment, achieving high performance in human preference tests.Meta · Nov 2025 · Unverified256
About this paper
Authors
Thaddäus Wiedemer, Yuxuan Li, Paul Vicol and 6 more
arXiv
2509.20328 · PDF
Venue
arXiv.org
Citations
215, 26 influential · Semantic Scholar
Upvotes
100 · Hugging Face
Lab
Google DeepMind · on Companies · on Acquisitions

Changes

What changed
Influential citationsfirst count: 26Sep 25, 2026
Citationsfirst count: 215Sep 25, 2026
New paperFound by the weekly scan, unverifiedSep 25, 2026

Sources: each lab's own papers and arXiv, with citation and upvote counts from Semantic Scholar and Hugging Face. One-line summaries are for orientation, not a substitute for the paper. Logos via logo.dev; trademarks belong to their owners.

New papers by email

Monday afternoons, only in weeks with new papers from the labs.

Double opt-in. Unsubscribe any time.