Reasoning effort switching without breaking cache is live in Codex and Claude
altryne · x · 2026-09-04
altryne notes that switching reasoning effort without breaking the prompt cache is now live in both Codex and Claude, via Astra and Fable. This lets developers dial reasoning levels up or down mid-session without invalidating the cache. He hopes the mechanism will also be extended to older models.
More from coding & agent
- Ethan Mollick: Treat Fable and Astra class models like an outside team, not an intern — emollick · 2026-09-04
- Your AI assistant only stores, never updates: a dev builds human-like memory with VectorAI — PrajwalTomar_ · 2026-09-04
- AI Won't Kill the Law Firm Apprenticeship Model—It Can Fix It — hoofnagle · 2026-09-04
- OUI-1: a fine-tuned Diffusion Gemma for Generative UI, 8x fewer params, open weights — GlennCameronjr · 2026-09-04
- GPT-6 Astra debuts at No.1 on Terminal-Bench, 1.9% ahead of Claude Fable 5.1 — sandersted · 2026-09-04
- Perplexity API lands in Stripe Projects: one CLI command provisions key and credits — jeff_weinstein · 2026-09-04