Models and development tools

Experiments and comparisons for choosing tools for a task, including the limits of each test.

4 articles

A cobalt-lit metal gate with a glass $500 plaque, while two blue light paths bypass the gate

Oct 20268 min read

OpenAI’s $500 Bet: Can Speed Justify the Premium?

OpenAI’s $500 Pro plan arrives as Claude and Gemini raise the bar. Can Ultrafast speed justify the premium for builders?

Eran Bodokh

Dark cover: the codename ox-alpha beside a glowing blue fingerprint dissolving into particles - the anonymous stealth model traced by its tokenizer

AI

Aug 20267 min read

ox-alpha: The Free Stealth Model on OpenRouter Is GLM-5.3-Flash

A free, anonymous frontier model called ox-alpha appeared on OpenRouter. We tested it the day it dropped. Z.ai has since confirmed it as GLM-5.3-Flash.

Eran Bodokh

A rising performance curve and abstract comparison panel illustrating Grok 4.6 price-performance economics

AI

Aug 202611 min read

Grok 4.6 Just Broke the Price of Frontier AI

Grok 4.6 matches GPT-5.6 Sol on an independent index at $2/$6 per million tokens. The benchmarks, caveats, Grok Bot, and Cursor deal.

Eran Bodokh

Claude Fable 5 vs GPT-5.6 Sol - parallel orchestration versus a modular research engine

AI

Jul 202613 min read

Claude Fable 5 vs GPT-5.6 Sol: Our Coding Verdict

Claude Fable 5 vs GPT-5.6 Sol for coding: benchmarks favor Sol, but our real-world tests still put Claude Code ahead in orchestration.

Eran Bodokh

Ideas, delivered to your inbox

Occasional practical guides and examples on AI and building products. No spam.