Models & Labs
Microsoft Unifies Copilot into Single 'Super App'
The Verge AIhigh confidence
Why it matters
- →Consolidates fragmented AI tools into a single enterprise interface.
- →Introduces usage-based billing for agentic capabilities like Autopilot.
- →Integrates GitHub Copilot coding features directly into the main Copilot app.
Microsoft has officially launched a redesigned Copilot application that consolidates its AI offerings into a single interface featuring Home, Code, and Autopilot tabs. The update rebrands the Scout personal assistant as Autopilot, an enterprise-grade agent with its own identity and cloud instance, while integrating GitHub Copilot’s coding tools directly into the workflow. Microsoft is introducing usage-based billing for these advanced features to manage costs associated with long-running agentic tasks. The rollout begins with the Frontier program for early access, with broader availability planned for later this year.
Read originalThe story around this
TopicMicrosoft Copilot Super AppCooling
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
More from The Verge AI
More in Models & Labs
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
llama.cpp Releases
Models & Labscoding
llama.cpp b12342 adds few-row MMQ for speculative decoding
Speculative decoding just got a significant speed boost on NVIDIA GPUs. This release introduces a specialized kernel for 'few-row' matrix multiplications, targeting the small batch sizes typical of verification passes in speculative decoding workflows. By optimizing these specific operations for quantized formats like Q8_0 and IQ4 variants, llama.cpp reduces latency where it matters most during token generation. The change is CUDA-only for now but leverages hardware capabilities from Ampere onwards, making local inference faster without requiring new model architectures or complex configuration changes.
