15 token-saving tricks for Claude: cut ~19,700 tokens per fix by scoping rewrites
goyalshaliniuk · x · 2026-09-18
A practical thread of 15 tricks to stretch Claude usage limits — the thesis: use it smarter, not harder or pricier.
- Convert files before uploading: a PDF page costs 1,500–3,000 tokens; the same content as markdown is under 200
- Stop "redo the whole thing": ask to redo only section 3 — saves 19,700 tokens per fix
- Edit your message instead of following up: every "No, I meant…" gets re-read in full
- Restart rather than keep replying: a 30-message session burns 232,000 tokens, mostly useless context
- New topic = new chat: Claude re-reads everything every turn
- Plan in Chat, build in Cowork: file creation costs more — think on the cheap tier
- Batch tasks into one message: one prompt = one context load
- Pick the right model: Sonnet/Haiku for simpler tasks (thread truncated here)
More from coding & agent
- Skip Scoping, Start Proving: A Five-Step Playbook for Shipping Enterprise AI at Startup Speed — alex_verem · 2026-09-18
- Berkeley study: the right agent harness cuts cost of the same result by 71% — MartinGTobias · 2026-09-18
- Meta Built an AI Agent as a Domain 'Secondary Expert' — With Auditable Knowledge Architecture — Meta_Engineers · 2026-09-18
- dotey on the AI code quality debate: black-box verification is replacing code review — dotey · 2026-09-18
- AgentSky launches as an agent marketplace: 40+ coding agents in browser, 44x cost gap between models — Scobleizer · 2026-09-18
- AWS Ships Six Open-Source Skills to Let Coding Agents Deploy Hugging Face Models on SageMaker — AWS ML Blog · 2026-09-18