Benchmarking a Real App Before and After Moving to the Edge
The empirical Edge performance audit. Benchmarking TTFB, Core Web Vitals (LCP/INP), global latency, and database query roundtrips before and after Edge migration.
8/1/202625 min read
#Benchmarking
6 articles tagged with Benchmarking
Benchmarking Claude, GPT, and Gemini on the Same Real Refactor Task
Head-to-head refactoring test. Claude Sonnet 5 vs GPT-5.6 Sol vs Gemini 3.5 Flash on a 3,000-line messy legacy TypeScript module. Code quality, architecture, and cost metrics.
8/1/202623 min read
I Benchmarked 5 AI Coding Agents on the Same Real Bug. Results Surprised Me
Head-to-head debugging benchmark. Claude Code, Cursor Composer, Devin Desktop, Windsurf, and Copilot Workspace tested on a complex async race condition.
8/1/202623 min read
The Real Latency Difference Between Flagship and Budget-Tier Models
Measuring Time-to-First-Token (TTFT) and tokens-per-second (TPS). A 2026 benchmark comparing GPT-5.6 Sol/Terra/Luna, Claude Sonnet 5 vs Haiku, and Gemini Flash.
8/1/202623 min read
Bun 1.2 vs Node.js 24 vs Deno 2.0: The 2026 Production Benchmark
The JavaScript runtime wars are over. Or are they? In this exhaustive 5,000-word benchmark, we test HTTP throughput, WebSocket latency, Cold Start times, and SQLite performance across the big three.
2/1/202626 min read
Next.js 14 Server Actions vs API Routes: Benchmarking Performance
Are Server Actions just hype? We benchmarked 10k requests on Vercel Edge. The results reveal a 40% reduction in cold starts and a massive shift in how we handle data mutations.
1/5/202614 min read