Search AI Products and News

Microsoft open-sources the Phi-4-mini-flash-reasoning model, specifically designed for edge devices, with inference efficiency improved by 10 times. It uses an innovative SambaY architecture to achieve efficient memory sharing, showing outstanding performance in long text generation and mathematical reasoning. Benchmark tests show its excellent long context understanding ability, with a Phonebook task accuracy rate of 78.13%. This model is suitable for educational and research fields and can run on a single GPU.

This morning, Microsoft officially announced its latest innovative small-parameter model, Mu. This model has only 330 million parameters, yet it can match the performance of Microsoft's previously released Phi-3.5-mini, while being just one-tenth the size of Phi-3.5-mini. More notably, Mu can achieve a response speed of over 100 tokens per second on offline NPU laptops, which is a rare breakthrough in the field of small-parameter models. A major highlight of the Mu model is its support for setting intelligent features in Windows.

Recently, the Central Cyberspace Administration launched the Clean Web Action - Special Campaign to Address the Abuse of AI Technology. Since April 2025, a series of rectification efforts have been carried out targeting the abuse of AI technology. The campaign focuses on addressing issues such as infringements on public rights caused by AI face swapping and voice imitation technologies, as well as misleading the public due to lack of content labeling. After the first phase, joint efforts from cyberspace administration departments at all levels have achieved remarkable results. Image source note: The image is generated by AI, and the image authorization service provider is Midjourney; during the rectification process, a cumulative total of

Significant progress has been made in the field of AI voice technology as Fish Audio announces the open sourcing of its new Text-to-Speech (TTS) model, OpenAudio S1-Mini. As a streamlined version of the highly-acclaimed S1 model, S1-Mini has triggered industry discussions due to its lightweight design, high expressiveness, and multi-language support. Key Features: Lightweight and High Performance OpenAudio S1-Mini is a lightweight version distilled from the 4B-parameter S1 model, containing only 0.5B parameters.

Recently, the Zcool Material Platform and WHEE have fully upgraded their AI functions and launched an innovative tool named "Mini Material Generator". The purpose is to help designers get rid of the cumbersome process of finding and processing materials, thereby significantly improving design efficiency.

On May 8th, OpenAI o4-mini launched with Reinforcement Fine-Tuning. The combination of these two technologies completely transforms the cost structure and technical threshold for AI specialization, enabling enterprises to rapidly turn general-purpose AI into expert systems in specific fields using only a small amount of training data. The leap from general intelligence to expert-level AI is highlighted by the Reinforcement Fine-Tuning technology at the core of this release. Unlike traditional supervised fine-tuning, RFT is based on reinforcement learning algorithms, driven by reward-based training.

OpenAI has announced the release of a free, lightweight version of its powerful AI research tool, Deep Research. This marks another significant step towards the democratization of AI technology. As an AI agent capable of independently completing complex research tasks, the free availability of Deep Research will provide students, researchers, and the general public with more convenient access to knowledge. Deep Research features: Intelligent Research Experience. Deep Research is an OpenAI...

OpenAI recently released its latest o3 and o4-mini AI models, which achieve state-of-the-art performance in many areas. However, these new models have not improved upon the issue of 'hallucinations,' exhibiting even more severe instances than previous OpenAI models. 'Hallucinations,' the generation of factually incorrect information by AI models, remain one of the most challenging problems in AI today. Previous generations of models showed improvements in reducing hallucinations; however, o3 and o4-mini have not.

xAI recently unveiled its new language model, Grok3Mini, further advancing efficient AI technology. Designed for speed and affordability, Grok3Mini, despite its smaller size, outperforms many more expensive AI models across various domains, particularly excelling in math, coding, and scientific benchmarks. Grok3Mini: The perfect balance of high performance and low cost. Grok3Mini is part of the Grok3 series, which includes six variants, including the standard Grok3.

With advancements in AI technology, OpenAI recently launched its latest inference models—o3 and o4-mini. These new models are not only more powerful in text understanding but also possess image reasoning capabilities, quickly becoming user favorites. According to TechCrunch, more users are leveraging ChatGPT to pinpoint the exact location where photos were taken, a phenomenon sparking widespread attention on social media. The o3 and o4-mini models offer powerful...

OpenAI recently launched a new system designed to monitor its latest AI reasoning models, o3 and o4-mini, to block prompts related to biological and chemical threats. The system aims to prevent the models from providing suggestions that could incite harmful attacks, ensuring AI safety. OpenAI states that o3 and o4-mini exhibit significantly improved capabilities compared to previous models, thus potentially posing new risks in the hands of malicious users. According to OpenAI's internal benchmarks, o3...

OpenAI recently unveiled its latest AI model, o3, marking a significant advancement in AI's ability to understand and analyze images, particularly low-quality sketches and diagrams. Alongside o3, a smaller version, o4-mini, was also released, expanding OpenAI's offerings. The core functionality of the o3 model lies in its ability to 'think with images,' allowing users to upload diverse images, such as whiteboard sketches and complex charts, for in-depth AI analysis and discussion.

OpenAI recently released its latest reasoning models, o3 and o4-mini, marking a significant breakthrough in the field of artificial intelligence. These models not only surpass previous versions in reasoning capabilities but also achieve image reasoning for the first time, integrating visual information directly into the thinking process. o3, hailed as a "genius-level" model, particularly excels in tasks such as programming and mathematics, achieving an accuracy rate of 87.5%. The newly released o3 and o4-mini models demonstrate exceptional performance in multimodal processing, possessing...

During a technical livestream at 1 AM today, OpenAI officially launched its latest and most powerful multimodal models: o4-mini and the full-power o3. These models offer unique advantages, capable of processing text, images, and audio simultaneously. They also function as agents, automatically utilizing tools such as web search, image generation, and code parsing. Furthermore, they possess a deep thinking mode, enabling reasoning about images within a chain of thought.

Opera, the Norwegian browser company, today announced that its AI assistant, Aria, is officially launching on the Android version of the Opera Mini browser. This move aims to make AI functionality easily accessible to users with low-end devices and limited data. Aria AI can help users quickly access the latest news, delve deeper into topics of interest, and even generate images. Opera stated that Aria's powerful capabilities stem from its integration with OpenAI and...

The competition in the AI field is heating up, and OpenAI is once again leading the charge with a technological breakthrough. AIbase learned from social media that OpenAI recently released three new models via API: GPT-4.1, GPT-4.1mini, and GPT-4.1nano. These models significantly outperform existing GPT-4 and GPT-4 mini models, particularly in coding, instruction following, and multimodal capabilities. This is AIbase's in-depth analysis of this significant update.