Alibaba Cloud’s Tongyi Qianwen has announced the Qwen‑MT machine translation model, which supports bidirectional translation across 92 languages
The machine translation model Qwen-MT, based on the Qwen3 model, has been officially released. It su...
Sorting
Qwen3 undergoes late-night upgrade, bidding farewell to hybrid thinking mode
Alibaba's Tongyi Qianwen has released the Qwen3-235B-A22B-Instruct-2507-FP8 version, featuring signi...
AI startup DecartAI has released MirageLSD, the world’s first real-time video generation model
AI startup DecartAI has unveiled MirageLSD, the world’s first live-stream video generation mod...
Princeton-led Team Releases the Most Powerful Open-Source Mathematical Theorem Proving Model: Goedel-Prover-V2
Princeton University, in collaboration with top institutions such as Tsinghua University, Peking Uni...
The release of Kimi K2 has sparked widespread discussion, surpassing many leading large language models.
Kimi K2, a MoE-based foundation model released by Chinese AI unicorn Moonshot AI, has garnered wides...
OpenAI’s acquisition of Windsurf has fallen through, as the company’s CEO and founder has been recruited by Google
OpenAI's acquisition of AI coding assistant Windsurf has officially been called off. Google moved qu...
The official announcement of collaboration between CAMEL-AI and Gemini to achieve data visualization and automation
The open-source multi-agent framework CAMEL-AI has announced a collaboration with Google DeepMind to...
The new image-to-video feature of Gemini
Gemini's newly launched image-to-video feature allows users to transform static images into videos w...
Musk releases Grok 4, capable of handling academic tasks at a PhD level
Elon Musk's xAI has unveiled its next-generation large language model, Grok 4, which demonstrates a ...
The Tongyi Lab has open-sourced its first audio generation model, ThinkSound.
The Tongyi Lab has open-sourced its first audio generation model, ThinkSound, specifically designed ...