Gemini 3 is Google's latest AI model, hailed as the world's most advanced multimodal understanding and reasoning model. The model boasts powerful reasoning capabilities, breaking numerous benchmark records, such as on the LMArena Leaderboard...
Generative UI is an innovative AI technology from Google that enables AI to instantly generate complete, interactive user interfaces based on users' natural language prompts. These interfaces can be web pages, tools, games, or educational applications, and are completely customizable...
MemOS is an open-source AI memory management operating system that provides long-term memory capabilities for intelligent agents and AI applications. MemOS supports various memory types (such as text-based memory, activation memory, parameter-based memory, etc.) through a unified memory abstraction layer.
Argus 1.0 is the world's first large-scale spatial model released by Ruishi that supports panoramic image input. Trained based on nearly one million sets of real high-definition spatial data accumulated by Ruishi, it adopts a Transformer architecture and features millisecond-level real-time reconstruction...
LoopTool is an automated, model-aware, iterative data evolution framework developed by Shanghai Jiao Tong University and the Xiaohongshu team. It improves the performance of Large Language Models (LLMs) in tool invocation tasks. The framework utilizes closed-loop optimization...
SAM 3D is an advanced 3D reconstruction model launched by Meta, comprising two sub-models: SAM 3D Objects and SAM 3D Body. SAM 3D Objects can reconstruct 3D models of objects and scenes from a single image, supporting multi-view...
GPT-5.1-Codex-Max is an intelligent programming model from OpenAI, designed for complex, long-cycle development tasks. Based on a newer inference architecture, the model uses a "compression" technique to span multiple context windows, enabling it to process large amounts of data...
SAM 3 (Segment Anything Model 3) is Meta AI's latest advanced computer vision model, capable of detecting, segmenting, and tracking objects in images and videos using text, examples, and visual cues.
PinMe is an open-source, one-click deployment tool that allows static websites or front-end projects to go live quickly. The tool uses IPFS distributed storage and ENS domains to generate permanent links, eliminating the need for servers and DNS.
SeekDB is OceanBase's open-source, AI-native database that addresses the pain points of multimodal data fusion and real-time processing in AI application development. It supports fusion queries combining vector retrieval, full-text search, and structured data filtering, all achievable with a single SQL query...
Nano Banana Pro (also known as Gemini 3 Pro Image) is a next-generation image generation and editing model from Google, incorporating Gemini 3 Pro technology. The model can generate high-quality, high-resolution (2K and 4K) images...
HunyuanVideo 1.5 is a lightweight video generation model open-sourced by Tencent's Hunyuan team, with a parameter size of 8.3B. Based on the Diffusion Transformer architecture, the model supports generating 5-10 second high-definition videos from text descriptions or images...
MiMo-Embodied is Xiaomi's first open-source, cross-domain embody model, integrating tasks from both autonomous driving and embody intelligence, achieving outstanding performance in environmental perception, task planning, spatial understanding, and many other aspects...
Supertonic is Supertone's open-source, high-performance text-to-speech (TTS) system, boasting extreme speed and lightweight design. Containing only 66M parameters, it generates speech up to 167 times faster than real-time, making it the fastest TTS system currently available...
Teammates is Ubisoft's AI teammate project based on generative AI technology, aiming to provide players with a more immersive and realistic gaming experience. Players can give real-time voice commands to their AI teammates, such as "Pablo, cover me" or "Sophie..."
WorldGen is Meta's latest advanced end-to-end system that generates interactive, navigable 3D worlds from a single text prompt. Users simply input a description, such as "medieval village" or "Mars base," and the system...
Olmo 3 is a series of large-scale open-source language models developed by the Allen Institute for Artificial Intelligence (AI2). The models include multiple versions, with Olmo 3-Base (the base model, with 7B and 32B parameters)...
Claude Opus 4.5 is Anthropic's latest advanced AI model. The model excels in programming, system-level tasks, routine research, and document processing, and is adept at handling complex software engineering problems.
Fara-7B is an open-source, agent-based Small Language Model (SLM) from Microsoft, focused on computer usage. The model helps users complete tasks such as filling out forms, searching for information, etc., by visually perceiving web pages and interacting with interface elements like the mouse and keyboard.
HunyuanOCR is an open-source, end-to-end OCR visual language model developed by Tencent's Hunyuan team. Based on Hunyuan's native multimodal architecture, it achieves state-of-the-art performance for multiple OCR tasks with only 1B parameters. It features a highly efficient and lightweight architecture...
FLUX.2 is a visual intelligence model from Black Forest Labs, designed specifically for real-world creative workflows. The model supports multi-image references of up to 10 images, generating high-quality images up to 4MP resolution, and boasts extremely high...
ViMax is an end-to-end multi-agent video generation framework developed by the Data Science Lab at the University of Hong Kong. It supports the automatic conversion of ideas, scripts, or novels into complete videos. The framework integrates the roles of director, screenwriter, producer, and video generator...
Z-Image is an image generation model launched by Alibaba Tongyi, featuring 6B parameters. The model includes three variants: Z-Image-Turbo, Z-Image-Base, and Z-Image-Edit, which excel in fast inference, basic development, and image editing, respectively...
Depth Anything 3 (DA3) is a visual space reconstruction model launched by ByteDance's Seed team. It uses a single Transformer architecture to recover the three-dimensional spatial geometry from visual input from any viewpoint.