Vercel Labs' open-source reference app for background coding agents.
Cloudflare has unified its multi-model AI gateway with BYOM support.
Qwen released a MoE model with 35B total parameters and 3B active parameters under Apache 2.0.
Mozilla has unveiled Thunderbolt, an open-source AI client focused on self-hosting and integration.
Cloudflare accelerated inference for massive LLMs through PD disaggregation, prompt caching, and KV-cache optimization.
Gemma 4 31B produced better results than competing models with fewer tokens in an image-based 3D geometry test.
pi-autoresearch extends Karpathy-style autonomous experimentation into a general-purpose optimization tool.
An account of Codex OAuth being blocked at `chatgpt.com/backend-api` due to Cloudflare bot protection.
AIOps Agent autonomously combines 7+ tools to cut incident analysis time by 93%.
The author argues that, facing the confusion and risks created by LLMs, we should pause and refuse.
Quint spins off from Informal Systems, positioning itself as an executable spec platform.
Built a tool that tracks AI agents' memory, decisions, and loops in real time using a 3D graph.
Cloudflare unveiled Artifacts, a version-controlled filesystem built for the agent era.
Cloudflare Workers now lets you create PlanetScale Postgres and MySQL directly, bundling billing together.
AI Search provides vector-keyword hybrid search and per-instance search.
Capcom's **PRAGMATA** arrives on **GeForce NOW** on its release day.
Turning Swagger into an SDK and harness makes AI build frontends more reliably.
llama.cpp's ggml added graph_reused to reduce CUDA's O(n) change checks.
Handle complex tasks with plan mode and step-by-step requests in Claude Code.
Global IPv6 access share has exceeded 50%, according to Google.
DFlash doesn't speed up as much as expected with MoE, quantization, concurrent use, and long context.
After enabling Firebase AI Logic, abnormal Gemini usage led to a €54k bill.
An image-to-SVG pipeline that runs the entire process locally, from Moondream to VTracer.