3 points | by dingdong2026 an hour ago
3 comments
OpenCode and cheaper plans await if you can move beyond the Big Ai labs and their bluster
Start auto-compacting your context at 100k. Your sessions will incur exponential token usage if you don't draw the line before the 1m context limit. It's the best way to efficiently use a small token quota for long-horizon tasks.
Um, you can put in $20 for more tokens any time.
OpenCode and cheaper plans await if you can move beyond the Big Ai labs and their bluster
Start auto-compacting your context at 100k. Your sessions will incur exponential token usage if you don't draw the line before the 1m context limit. It's the best way to efficiently use a small token quota for long-horizon tasks.
Um, you can put in $20 for more tokens any time.