User tests how Claude Code, Cursor and Codex bill resumed sessions
11/18/2026 — 11/18, 17:19·1 sources (all)·1 reports (all)
Key takeaways
- After pushing a context to 546,151 tokens, one user compared how Claude Code, Codex, Antigravity and Cursor bill resumed sessions. The worki…
Background & analysis
On November 18, 2026 at 17:00, the AI section of DEV Community published a hands-on article asking how much it costs to resume an old session and how much the cache really saves. A user named Anton, after pushing his own context to 546,151 tokens, set out to compare how Claude Code, Codex, Antigravity, and Cursor handle billing when an old session is resumed.
His working hypothesis was that if you step away for longer than 40-60 minutes and then return, the entire context history has to be sent again and every token is paid for a second time, while staying inside the cache window keeps costs lower because the cache takes effect. The article, however, argues that the real cache window may be just 5 minutes rather than the commonly cited 40-60 minutes, and stresses that what ultimately determines the cost of resuming an old session is the user's own context, which in Anton's test had reached 546,151 tokens.
At this point the report only lays out the testing plan and the assumptions behind it. It has not yet produced a final, side-by-side breakdown of what resuming an old session costs in each of the four tools, so the comparison remains unresolved and no definitive conclusion is drawn.
AI-generated from 1 reports · updated 4 days ago
Related AI tools
- Heat index
- 25
- All sources
- 1
- All reports
- 1
- First seen
- 4 days ago
Reports on this story headlines open the original
After his context grew to 546,151 tokens, one user set out to compare how Claude Code, Codex, Antigravity and Cursor charge for resuming an old session. His hypothesis: past the cache window, the whole history is resent and paid for again, and that window may be 5 minutes rather than the 40-60 often assumed.
DEV Community · AIAI score 54
Other stories people are talking about
- Heat index 32Local-first MCP memory for Claude Code and Cursor, no vector DB1 sources
- Heat index 63Developer benchmarks Claude Code Code Mode: 44% cheaper despite more requests1 sources
- Heat index 34Third-party guide compares Claude Code subscription tiers and API pricing1 sources
- Heat index 38Developer ports 908 VS Code themes to Claude Code1 sources
- Heat index 75Developer breaks a Claude Code mod five ways to test validation1 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

