artifact

Not waiting for GPT-5.3, Gemini 3.5, this week's big model release wave was first by Ali “jumped the gun”! Last night, Ali launched Qwen3-Max-Thinking, which is the flagship reasoning model with the strongest capability of Ali's Thousand Questions series, with 19 authoritative benchmarks...

Recently, at the Wenxin Moment conference, the official version of Wenxin Big Model 5.0 was launched. It is said that the model parameter count reaches 2.4 trillion, adopts native full-modal unified modeling technology, has full-modal comprehension and generation capability, and supports input and output of text, image, audio, video and other kinds of information. ...

On December 11, Apple published a paper introducing the 3D generation model SHARP, claiming that on standard GPUs, the model can reconstruct a single image into a photorealistic 3D scene in under one second. The model is now open-source. Users need only input a regular photo, and the model can generate a 3D scene through neural...

Just now, Google unveiled its latest model, Gemini 3 Flash. According to the introduction, it boasts cutting-edge intelligence and is built for speed, empowering everyone to learn, build, and plan anything faster. Even before the model's release, the Google AI Studio developer platform...

In the early hours of today, coinciding with OpenAI's tenth anniversary, the company officially launched its most powerful model to date, GPT-5.2, alongside the simultaneous release of ChatGPT and its API system. This update includes three variants of GPT-5.2: Instant, Thinking, and Pro...

“Lingguang”, completely on fire. Last week, Ant Group released the full-modal universal AI assistant Lingguang, perhaps one of the biggest surprises in the AI application track at the end of 2025. Lingguang's downloads showed accelerated growth, and on the first day of the launch, Lingguang's downloads exceeded 200,000, and on the fourth day exceeded 100...

Early this morning, Google's strongest reasoning model, Gemini 3, was finally unveiled, a model that encompasses a wide range of native multimodal, reasoning, and Agent capabilities. The Google DeepMind research team says it is the world's most advanced multimodal understanding model, Google's most powerful Agent programming and atmosphere...

Since the beginning of this year, the open source big model battlefield has been full of smoke. People from all walks of life are fully invested here, trying to seize the ecological position in the next era of AI. And a trend that can't be ignored is that Chinese big models, are strongly ruling the 'Hall of Fame' of open source base models. From DeepSeek in Code and Math...

October 23, in tencent AI workbench ima released the first anniversary of the event site, ima2.0 version of the updated content was officially announced, including the Agent ability based on the “task mode”, as well as the knowledge base “AI key points” and other functions, will be launched on the 24th internal beta test on the 27th. It will be launched on the 24th for internal testing, and will be officially launched on the 27th...

On October 21st, local time, OpenAI officially launched ChatGPT Atlas, an independent AI-powered web browser, through an online live broadcast.This ChatGPT-centered browser is not only the first time for OpenAI to venture into the field of Internet terminal tools, but also the first time for users to access the web...

Baidu open source multilingual document parsing model PaddleOCR-VL on October 16, three consecutive days to dominate the Hugging Face Trend List first. PaddleOCR-VL can recognize complex elements such as text, tables, formulas and charts in 109 languages, including all...

Today, Google announced the launch of the new Veo 3.1 video generation model, the new version of the improved picture realism and narrative control, while having a more powerful cue word comprehension and audio-video synchronization generation capabilities. Relying on the capabilities of Veo 3.1, Google has simultaneously upgraded its AI film and television creation platform...

In the early hours of October 1, OpenAI released its flagship video and audio generation model, Sora 2, and launched the "AI version of Jitterbug" - the Sora app. OpenAI calls Sora 2 "a GPT-3.5 moment straight out of the video space". Compared to the previous Sor...

Today, Claude Sonnet 4.5 was officially released, a model that can sustain focused processing of complex multi-step tasks for more than 30 hours, topping the list of programming skills, computer manipulation skills, and surpassing the GPT-5 in many aspects of reasoning, math, and programming of intelligences. Claud...

September 26, Tencent announced that the hybrid 3D generation model family added new members - hybrid 3D-Omni, hybrid 3D-Part released and open source. This is a new breakthrough for Tencent's hybrid in controllable 3D generation, making AI 3D modeling more practical and accelerating the 3D generation model in the tour...

The Large Language Model (LLM), is entering the Next Level. In the early morning of Friday, Ali Tongyi team officially released and open-sourced the next-generation base model architecture Qwen3-Next. 80B of total parameters of the model is only activated 3B, the performance can be comparable to the flagship version of the Qianqian 3 235B model...

On September 8, Ali released its latest speech recognition model Qwen3-ASR-Flash, which is trained based on the Qwen3 base model and supports 11 languages and multiple accents. Users can use ModelScope, HuggingFace and AliCloud Hundred Refined API...

On September 5, Ali went live with the Preview version of Qwen3-Max, the strongest model in the Qwen3 series, which is also Ali's largest model to date, with more than 1 trillion participants. The model has now landed on the Ali Bailian platform, and is available for free on the Tongyi Thousand Questions app and Qwen Chat...

On August 25, Nail held a conference to release version 8.0, marking the largest user scale of the mobile Internet era of collaborative office APP officially to AI native. Nail CEO Wuzhao said: "Nail 8.0 is also the 1.0 of AI Nail, we are determined to empty the past, with a zero mentality, for A...

Enterprise WeChat 5.0 is officially released, the new version focuses on updating the intelligent search, intelligent summary, intelligent robot, form collaboration, service record organization, enterprise business card and other functions, but also in the mailbox, internationalization and private deployment of the expansion. First, AI helps office efficiency Intelligent summary The work of...