Miles Brundage: Codex Ultra Fast burns budget in hours, widening the compute gap
Miles_Brundage · x · 2026-10-01
Former OpenAI policy lead Miles Brundage shares his experience with Codex: he almost never exhausted his Codex budget before (with Claude it happens more often), but the new Ultra Fast mode is so fast he can burn through it in a few hours, leaving him feeling compute-poor.
He argues developers now fall into roughly three compute tiers: free tier, paid tier, and employees at a handful of companies — and the gap between the second and third tiers has widened significantly with multi-agent workflows and now raw speed.
Related event: Ex-OpenAI exec feels "compute-poor" after trying Codex Ultra Fast(2 posts)→
More from coding & agent
- Dev Builds a Minimal Coding Agent From Scratch, With Session Branching and Compaction — nik-55 · 2026-10-01
- Cognition First to Run NVIDIA Vera Rubin on CoreWeave, ~4.8x Throughput vs GB200 — silasalberti · 2026-10-01
- Astra ultrafast rebuilds an app from scratch in ~3 mins using 10% of a $500/mo plan — OpenAIDevs · 2026-10-01
- ChatGPT's 'Dot' Can Spin Up Its Own Threads and Message Them Proactively — daniel_mac8 · 2026-10-01
- Developer finally vibe codes his dream ambient music DAW after years of ideating — kieranklaassen · 2026-10-01
- Dots now tap your Codex and ChatGPT context and can run multiple tasks at once — dkundel · 2026-10-01