VideoThe Gray Cat · Sep 1, 2026
Caveman vs Ponytail vs RTK: Which Saves Most Tokens?
I tested Caveman against Ponytail, then added RTK to the Claude Code comparison. Ponytail used 61% fewer tokens than plain Opus 5. The feature added custom vocabulary hints to transcription. Every…
More to watch
The Gray Cat · Oct 27, 2026
OpenCode Go + CodeGraph: Can Cheap Models Do the Job?
Open harnesses · Cost & limits
The Gray Cat · Oct 20, 2026
I Tested 4 Codex Frameworks on the Same Real Feature
Codex & OpenAI agents · Comparisons & reviews
The Gray Cat · Oct 13, 2026
Understand Anything vs Groma: Two Ways to Learn Your Code
Context engineering & memory · Comparisons & reviews
AI Engineer · Oct 29, 2026
How Revolut Cut AI Coding Costs 90% With Self-Hosted Open Models - David Abitbol, Revolut
Cost & limits · Open harnesses
AI LABS · Oct 28, 2026
This Is Microsoft's New Plan To Fix Claude's Usage Limits
Cost & limits · Claude Code
Theo - t3.gg · Oct 27, 2026
the hard truth about token spend
Cost & limits