Power user burns 1bn tokens a day at 98.65% cache hits, weighing GLM-5.3 vs DeepSeek after Claude Code weekly limits
julianharris · x · 2026-09-19
A heavy user reports consuming 1bn tokens in a single day (98.65% cache hit rate) while running 3-4 coding sessions across 2-3 projects, and is about to hit Claude Code's weekly limits. He is evaluating GLM-5.3 or DeepSeek 4.1 as alternatives, noting the $20 trial pack's 60m tokens would burn through in minutes at his usage rate.
More from coding & agent
- OpenClaw ships official WhatsApp channel plugin on the Agent Platform API, open-source — heyneighbor · 2026-09-19
- Five tools for running Claude Code and Codex agents in sync — Alert_Hunt_3259 · 2026-09-19
- TipTour goes open-source: local CoreML-driven Mac pointer agent clicks in ~90ms without screenshots — alexcovo_eth · 2026-09-19
- Dev hides AI thinking notes to stop spoilers as GLM-5.3 Flash stays on cost Pareto frontier in chess app — MikePFrank · 2026-09-19
- An AI agent called customer service for a user — the rep couldn't tell it wasn't human — armand_ruiz · 2026-09-19
- GuppyLM: Train a 9M-parameter LLM from scratch in 5 minutes with one Colab notebook — tom_doerr · 2026-09-19