Alibaba launches Qwen3-Max with trillion-parameter scale and adds open-source Qwen3-Omni, a real-time multimodal model with competitive benchmarks and clear API pricing.
Microsoft added Anthropic’s Claude models to Microsoft 365 Copilot—starting with Researcher and Copilot Studio—broadening model choice beyond OpenAI
Google has introduced Mixboard, an AI-powered mood board app available in U.S. beta, designed to rival Pinterest and Canva with AI-driven creativity.
Luma AI’s Ray3 debuts with Adobe Firefly integration, enabling reasoning-based video generation, HDR EXR output, and creative workflows in Creative Cloud.
Gemini 2.5 Deep Think earned ICPC 2025 gold in Baku, solving 10/12 problems—including one no human solved—in under 30 minutes, with a 677-minute combined time.
Seedream 4.0 merges generation and editing, adds six-image reference consistency and natural-language edits, delivers 2K outputs in <2s, and leads blind arenas.
CAS unveils SpikingBrain1.0, a brain-inspired LLM trained on MetaX GPUs with hybrid-linear attention, reporting 100× TTFT at 4M tokens and high sparsity for low power.
Built by EPFL, ETH Zurich, and CSCS, Apertus is a fully open LLM trained on 15T tokens across 1,000+ languages, with EU-aligned transparency and access via Swisscom and Public AI.
Microsoft introduces MAI-Voice-1 and MAI-1-preview, a mixture-of-experts model trained on ~15,000 Nvidia H100s, with GB200 compute active and targeted Copilot rollout.
