AI Product Updates: New Features and Releases

Qoder has moved beyond its web version with a desktop app, and its site and startup onboarding are praised as top-tier domestic design.

Admins can send Copilot agent activity data to their organizations' monitoring tools to analyze sessions and investigate issues.

Xiaomi's open-weights MiMo-V2.6-Pro and Flash multimodal reasoning models are now on B.AI, with a 1M context window and text, image, video and audio understanding.

Video generation platform Palmier Pro has launched Anthropic's new Opus 5.5 model and supports calling it through MCP.

The Claude mobile app now lets users sign in to multiple accounts at once for easier switching.

Higgsfield announced integration of Anthropic's new Claude Opus 5.5, which it says is 40% cheaper and over 30% faster than Opus 5.

Alibaba's Qwen says Qwen-Image-2.1 ranks first among open-source models in both the Arena image editing and text-to-image leaderboards.

Firecrawl closed a $75M Series B led by Smash Capital and released Alexandria, which it bills as the world's largest knowledge library.

Higgsfield used Claude Opus 5.5 to turn Monet's paintings into a walkable 3D world, with models, animation, music and sound all generated inside its platform.

The open-source pd-bridge project runs prefill on DGX Sparks and decode on Mac Studios, moving the KV cache between them over RDMA via MCDMA.

Building on its AI sleep coach with 500,000 registered and 80,000 daily active users, Gotchi Labs is expanding into shopping, productivity and personal finance, with a dedicated agent per vertical and $CHI as the shared layer.

SideShift released three products: an Agent that runs ad campaigns end to end, Commission Ads paid through a share of ad spend or sales, and Audience Data for verifying where creator videos come from.