28d • 🚨 Help & Troubleshooting
Tips to reduce cached tokens in Claude Code?
Hey everyone,
Looking for your best tips to optimize token usage in Claude Code.
My issue: I have a fairly heavy file stack ( multiple CLAUDE.md files, context files, auto-memory, etc.). Every new conversation starts with a large amount of cached tokens loaded before I even type my first message. I hit limits fast and can't work efficiently on longer tasks anymore.
Would love to know:
- How do you manage your context files to avoid loading everything each time?
- Ways to structure a project so Claude doesn't load everything unnecessarily?
for exemple : Cached: 54.9k | In: 954 | Out: 10.5k | Total: 54.9k -> after 1 message
Thanks in advance!
13
26 comments
Tips to reduce cached tokens in Claude Code?
skool.com/quantum-quill-lyceum-1116
Jake Van Clief, giving you the Cliff notes on the new AI age.
Leaderboard (30-day)
1
+695
2
+621
3
+423
4
+414
5
+288
Powered by