Articles, one at a time.
Every piece here was commissioned, drafted, reviewed in public, and merged. No content mills, no auto-published slop.
On-Device AI 2026: Developer Guide to NPUs and Edge Inference
A practical 2026 guide to on-device AI: NPU vs GPU vs CPU for LLM inference, Apple M5 MLX, Qualcomm X Elite, Core AI for iOS 27, and edge deployment.
Read →
DeepSeek V4-Pro and V4-Flash: Migration Guide and API Setup
DeepSeek V4-Pro (1.6T MoE, 1M context) and V4-Flash released April 2026. Migrate before the July 24 deadline. Full API guide, benchmarks, pricing.
Read →
Meta Llama Stack: Deploy Llama 4 With OpenAI-Compatible API
Deploy Llama 4 to production with Meta Llama Stack's OpenAI-compatible API. Covers distributions, vLLM, Ollama, safety, agents, and cost-effective hosting.
Read →
nanobot: Build AI Agents in 4,000 Lines You Can Actually Read
nanobot is a ~4,000-line Python AI agent by HKUDS. Connect 8+ platforms, 11+ LLM providers, and read the entire source in an afternoon.
Read →
Cursor 2.0: 8 Parallel AI Agents and Visual Editor Bridge
Cursor 2.0 ships Composer, up to 8 parallel agents, and a visual editor bridge. Full review of features, pricing, and workflow for developers in 2026.
Read →
GPT-5.5 Spud: Unified Multimodal API — Developer Integration Guide
GPT-5.5 Spud is OpenAI's first natively omnimodal model. One API call handles text, audio, image, and video. Here's how to use it as a developer.
Read →
Llama 4 Maverick: 400B MoE Model — Self-Hosting and API Guide
Complete developer guide to Llama 4 Maverick: MoE architecture, hardware requirements, vLLM setup, API providers, and benchmarks vs GPT-4o.
Read →
Databricks Unity AI Gateway: MCP Agent Governance Guide
Learn how Databricks Unity AI Gateway governs MCP agents with fine-grained permissions, LLM safeguards, and end-to-end observability.
Read →
GitLab 18.11: Agentic AI for Security, CI, and Analytics
GitLab 18.11 ships three agentic AI features: SAST auto-remediation, CI Expert Agent, and Data Analyst Agent. What developers need to know.
Read →
Kimi Code K2.6: Moonshot AI's Coding Model vs Claude Code
Kimi Code K2.6 review: 58.6% SWE-Bench Pro, 300-agent swarms, $0.60/M input. How it compares to Claude Code in real-world coding tasks.
Read →
LLM Inference Engines Compared 2026: vLLM vs SGLang vs TGI vs MAX
Compare the top LLM inference engines in 2026: vLLM, SGLang, TGI, and MAX. Real benchmarks, architecture deep-dives, and which to pick for production.
Read →
Qwen3.6-Plus: 1M Token Context and Claude-Level Performance
Alibaba's Qwen3.6-Plus packs a 1M token context, agentic coding, and hybrid MoE architecture — at 18x lower cost than Claude Opus 4.6. Developer guide.
Read →