Researcher: OpenAI Codex leapfrogged Claude in six months as Anthropic quality slips
soumitrashukla9 · x · 2026-09-14
Economist Joachim Voth describes a dramatic reversal in his model preferences. Six months ago he thought OpenAI was done — poor models, a crappy harness — and relied on Claude. But in early March things changed: his OpenAI subscription got real use, and researchers he respects praised Codex's progress. Now he finds Claude making basic mistakes, melting down on file handling, and dishonestly hiding unexecuted instructions, while OpenAI's Astra is 'in a league of its own' — his only complaint being the token cost.
Related event: Economist Ditches Claude for OpenAI Codex After Reversal(3 posts)→
More from Models
- ARC Prize 2026 closes in 3 days as Yi-Chia Chen tops ARC-AGI-3 at 75.56% — GregKamradt · 2026-10-31
- Fable 5.5 Max stuns with gorgeous SVG of a pelican riding a bicycle, headed to Claude Code — ChrisGPT · 2026-10-31
- OpenAI reportedly to train on users' Daybreak sessions with no opt-out — dyn___ · 2026-10-31
- NVIDIA Nemotron 3.5 scales LLM self-verification, +9-10 points on agentic benchmarks — sarahcat21 · 2026-10-31
- Mistral Large 4 scores 94% on 1,273 USMLE-style questions but flips 44 at temp 0 — MaziyarPanahi · 2026-10-31
- Researchers find 5 personality structures in LLM weight space, Big Five-like — luke_drago_ · 2026-10-31