Pydantic AI Review 2026: Type-Safe Agents in v2, Tested Without an API Key
Pydantic AI review: v2.68 run offline with real output. Validation retries, docstring tool schemas and hooks work well, but importing it alone takes 1.1 s.
Pydantic AI review: v2.68 run offline with real output. Validation retries, docstring tool schemas and hooks work well, but importing it alone takes 1.1 s.
Vibe coding in 2026, from someone who ships vibe-coded mini-apps: where it saves real time, where it quietly breaks in production, and an honest verdict.
Django-Bolt puts Actix and Rust under Django to hit 188K RPS. I dug through the source and benchmarks — where it beats FastAPI, and where the speed evaporates.
Claude Sonnet 5 review: $2/$10 is now its permanent price, the tokenizer adds up to 35% more tokens, and Sonnet 5.5 costs the same. Which one to run now.
Bumblebee scans npm, PyPI, Go, MCP configs and editor extensions for compromised packages without running install scripts. What it catches, where it fits.
GLM-5.2 scores 62.1 on SWE-bench Pro vs GPT-5.5's 58.6, ships under MIT, and costs $1.40/M input tokens. Benchmarks, pricing, and the China data question.
Claude Fable 5 scored 80.3% on SWE-bench Pro at $10/$50 per 1M tokens. What changed since: Fable 5.1, then Fable 5.2 at $8/$40, opt-in fallback and ZDR.
Devin Desktop is Windsurf rebranded by Cognition: an Agent Command Center, Devin Local in place of Cascade, ACP support, and Free, $20 and $200 plans.
A GitHub Copilot AI credit costs $0.01: 8,333 output tokens on GPT-5.6 Luna, 500 on Sol. Per-model token math, plan credit pools, and the Oct 19 model swap.
Google Jules went read-only on Nov 11, 2026 and shuts down Dec 31. What the async coding agent did well, how to export your data, and which agents replace it.