AMD Ryzen AI Halo: Run 200B Models, Drop the API Bills
AMD's $3,999 Ryzen AI Halo mini-PC runs 200B-parameter models on-device. Here is what developers need to know about specs, costs, and Perplexity's local ...
Aspire 13.6: Persistent Dashboard, REPLs, Java & Rust
Aspire 13.6 ships with SQLite-backed dashboard history, one-click database REPLs, and first-party Java and Rust ...
GitHub Copilot HydraFusion: Multi-Model Routing, 67% Cost Cut
GitHub's HydraFusion routes each Copilot coding task through Single, Cascade, or Critique workflows automatically — ...
Google EnvHarness: Open-Source Agent Training That Adapts to Your AI
Google open-sourced EnvHarness on Oct 2: a wrapper that reshapes agent training benchmarks without touching ...
Turbo Haskell: Edward Kmett Brings Haskell to the JVM
Edward Kmett shipped a JIT compiler for Haskell on GraalVM/Truffle in about a week. THC ...
Amazon Strands Decider 2B: Agent Decisions, No LLM
Amazon's open-source Strands Decider 2B makes agent routing decisions in 106ms without generating text. Here's how it works, where it fits, and the ...
MAI-Transcribe-2-Streaming: Real-Time Voice Agents, Completed
Microsoft launched MAI-Transcribe-2-Streaming on October 1 — a streaming STT model ranking #1 across 38 ...
Antigravity 2.0 on Googlebook: Your New Dev Workflow
Googlebook OS ships October 4. Here is what developers need to know about Antigravity 2.0 ...
GPT-6.1 Sol Is Out: Near-Astra Performance at One-Fifth the Price
GPT-6.1 Sol delivers near-Astra performance at $2/$10 per million tokens—5x cheaper than Astra. What changed ...
Chrome DevTools for Agents: Third-Party Tools API
Chrome 153 and 154 add a third-party tools API to DevTools for Agents, letting frameworks ...