Google updates its specialized Gemini 3 'Deep Think' reasoning mode to better solve modern science, research, and engineering challenges and expands access via the Gemini API to some researchers.
Seedance 2.0 goes viral in China, gaining rapid public attention and usage.
OpenAI introduces GPT-5.3-Codex-Spark, an ultra-fast model for real-time coding (~1,000 tokens/s) released as a research preview.
Seedance 2.0 is described as production-ready and supports 15-second multi-shot outputs with dual-channel audio.
YouTube launches an official app for the Apple Vision Pro, available starting today.
MiniMax releases the MiniMax-M2.5 model, claiming it delivers on the 'intelligence too cheap to meter' promise.
Zhipu AI releases GLM-5, an open-source LLM with enhanced coding capabilities and support for long-running agent tasks.
OpenAI releases GPT-5.3-Codex-Spark, its first AI model to run on Cerebras Systems chips.
Ever positions itself as the first "AI-native, full-stack" retail marketplace for electric vehicles.