OmAI Lab (@OmAI_lab) header
OmAI Lab (@OmAI_lab) profile picture

VLX — Edge-native multimodal AI for real-time perception, reasoning, and action. GitHub: https://teapawt.net/t.co/ZUYXqCM9Lj Discrod: https://teapawt.net/t.co/NJjOYSxgav

Joined October 2024
Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)Photo by OmAI Lab (@OmAI_lab)
(1/4) Excited to share our latest work from Om AI Research and ZJU! As we push toward Vision-First architectures for physical AI, a critical question remains: Which pre-training method provides the best foundation model for Spatial Intelligence? VLM or VGM? 🧵👇
1
3
4
459
Latest blog on teaching VLMs to understand fine-grained objects. VLM-FO-1 equips a novel object-enhanced vision tower, achieving remarkable object understanding performance with only 3B parameters. Larger models and RL-enhanced coming soon. om-ai-lab.github.io/2025_08_…
2
2
12
1,212
1/3: 🚀 Thrilled to share VLM-R1’s latest results! After hitting SoTA in REC & Math, we’ve supercharged RL for open vocab detection (OVD). TL;DR: With the right rewards, RL-powered VLM nails SoTA on OVD + sparks cool "aha" moments. Dive in: om-ai-lab.github.io/2025_03_…
2
4
2
12
1,365
Introducing VLM-R1! GRPO has helped DeepSeek R1 to learn reasoning. Can it also help VLMs perform stronger for general computer vision tasks? Our preliminary answer is YES and it generalizes better than SFT. github.com/om-ai-lab/VLM-R1
6
94
7
560
58,564
OmAI Lab retweeted
OPENAI ROADMAP UPDATE FOR GPT-4.5 and GPT-5: We want to do a better job of sharing our intended roadmap, and a much better job simplifying our product offerings. We want AI to “just work” for you; we realize how complicated our model and product offerings have gotten. We hate the model picker as much as you do and want to return to magic unified intelligence. We will next ship GPT-4.5, the model we called Orion internally, as our last non-chain-of-thought model. After that, a top goal for us is to unify o-series models and GPT-series models by creating systems that can use all our tools, know when to think for a long time or not, and generally be useful for a very wide range of tasks. In both ChatGPT and our API, we will release GPT-5 as a system that integrates a lot of our technology, including o3. We will no longer ship o3 as a standalone model. The free tier of ChatGPT will get unlimited chat access to GPT-5 at the standard intelligence setting (!!), subject to abuse thresholds. Plus subscribers will be able to run GPT-5 at a higher level of intelligence, and Pro subscribers will be able to run GPT-5 at an even higher level of intelligence. These models will incorporate voice, canvas, search, deep research, and more.
3,587
4,141
2,060
36,974
7,147,199
In EMNLP, I attended a great agent tutorial hosted by @ysu_nlp @Diyi_Yang @ShunyuYao12 , learned a lot about language agents and the progress in agentic ops, e.g. ReAct, CoT etc. How do these methods perform given the same model? I did a study to find out! github.com/om-ai-lab/open-ag…
1
7
26
5,001
OmAI Lab retweeted
I just created this cool RPG text-based game with a dynamic storyline that adapts based on your choices — and it only took me 5 minutes! ⏳ While it looks complex, OmAgent made it incredibly easy to build and customize. Best part? You can connect it to your phone and interact with your own creations! 📱github.com/om-ai-lab/OmAgent #AI #AIAgent #OmAgent #opensource
2
7
892
OmAI Lab retweeted
🚀 OmAgent now supports calling DeepSeek r1 through @ollama! This means you can integrate DeepSeek’s incredible model into your workflows seamlessly using OmAgent. 🎉Thanks to @deepseek_ai for training such an excellent model at such a low cost, and it's completely open-source! 🙌Try it out with OmAgent today and see how easy it is to get started! 🔥 #OpenSource #LLMs #MachineLearning #AI #AIAgent
2
7
455
Came across some amazing work by @OmAI_lab! 🔥 ✨ Open Agent Leaderboard: Track the best AI agents in one place! huggingface.co/spaces/omlab/… ✨ OmAgent: A multimodal agent framework for video understanding, handles everything from CCTV to full-length films 🎥 huggingface.co/papers/2406.1…
3
5
1
12
1,266
OmAI Lab retweeted
【OmAgent v0.2.2 Now Officially Release!】 This update brings comprehensive improvements, significantly enhancing the development and application experience. 👇 🌟 New Features: Introduced intelligent algorithms and a comprehensive intelligent agent evaluation platform. -Agent Operator: Developers can now seamlessly integrate these operators into their projects to implement complex intelligent functionalities. -Open Agent Leaderboard: The first open platform for evaluating intelligent agent algorithms. 🚀 Enhancements: -Optimized prompt injection, supporting variable transmission through configuration files. -Improved worker scheduling mechanisms. -Optimized module registration, prohibiting duplicate registrations. -Provided and tested local large model deployment solutions with Ollama. Welcome to experience the new features now! Github Link:github.com/om-ai-lab/OmAgent… #Ollama #AI #OpenSource #OmAgent
1
10
365
OmAgent - 简单易用的多模态 AI 智能体开发框架 概述: @OmAI_lab 开源的一个让开发者能轻松构建和部署能理解文字、图像、视频、音频的 AI 应用的开源框架,通过封装复杂的底层技术,实现了像搭积木一样简单地开发多模态 AI 应用 主要特色功能 · 提供灵活的智能体架构,支持图形化的工作流编排 · 原生支持多模态交互,包括视觉语言模型(VLM)、实时 API 和计算机视觉模型 · 内置多种先进的推理算法,如 ReAct、CoT(思维链)等 实用案例 · 视频问答系统:能够理解和回答关于上传视频的问题 · 移动个人助手:可以快速构建类似 Google Astral 的多模态助手 · 支持多种推理方式的对比实验,如 SC-COT、ReAct-Pro 等算法的性能对比
OmAgent python library for building multimodal language agents
2
1
4
1,255
It's only been 1 day since CES 2025, and people are going crazy it. This will change the way we use technology forever. 13 most jaw-dropping reveals so far (Don't miss the 5th one)
264
2,318
210
28,341
9,369,347
OmAI Lab retweeted
NVIDIA just broke the Internet at CES2025. Jensen Huang dropped some mind blowing announcements. Here are 10 key reveals you can't miss:
144
952
98
8,822
2,182,521
Sam Altman, in a new blog post, says "we are now confident we know how to build AGI" and OpenAI are "beginning to turn our aim beyond that, to superintelligence"
128
246
119
1,801
350,459
OmAI Lab retweeted
Finding specific items in retail stores has never been easier! With OmAgent's real-time video analysis, we quickly detect the exact location of Coca-Cola in the fridge. This technology not only enhances efficiency but also empowers visually impaired individuals by guiding them to their desired items.This isn't just image recognition—it's a dynamic solution that adapts to real-world environments, revolutionizing the retail experience! #SmartRetail #VideoAnalysis #RealTimeTracking #OmAgent #OmAI #AgentFramework #AI
2
2
8
460
OmAI Lab retweeted
🎉 Excited to launch my new course, Introduction to AI Agents! Every AI dev and company I’ve worked with this year is keen to build with AI Agents. I’ve worked hard to put all my learnings into this course so you too can learn how to build agentic AI systems. Material ranges from fundamentals to practical tips to help you build advanced agentic workflows. Enroll now: dair-ai.thinkific.com/course… This course is for you if you are looking to apply AI agents in a professional setting. No programming is required for this course! Use code AGENTS20 to get an extra 20% discount now. (Prices will increase soon so make sure to take advantage of the current offer)
25
100
5
719
141,200