Burkov's post-update workflow: fast tasks vs slow tasks now split across differently named model tiers
burkov · x · 2026-10-06
ML author Andriy Burkov shares his workflow after recent GPT and Claude updates: for quick turnarounds he uses the "Extra"-tier Sonnet-style model, while tasks that can wait until tomorrow go to the "High"-tier slow-and-thorough option. He notes this mirrors his old setup — Codex for speed, OpenCode with open-weight models for anything that could wait — but the roles have now flipped to official models. The post also doubles as a jab at the increasingly confusing model naming schemes across vendors.
More from coding & agent
- OpenCode hits 17M MAUs just 16 months after launch — ycombinator · 2026-10-06
- Claude Code 2.1.291 Fixes Session-Drop Regressions, System Prompt Tokens Jump 15% — ClaudeCodeLog · 2026-10-06
- Claude Code 2.1.291 fixes permission-prompt and message-loss regressions — ClaudeCodeLog · 2026-10-06
- Claude Code 2.1.291 fixes cloud session and message-loss regressions — ClaudeCodeLog · 2026-10-06
- AI agent architecture explained: the 7 modules from perception to observability — ZabihullahAtal · 2026-10-06
- Memoria 1.0: a local, model-agnostic LLM memory layer hitting 89.8% Recall@1 on LongMemEval — kitkatz69 · 2026-10-06