34 lines
1.3 KiB
Markdown
34 lines
1.3 KiB
Markdown
---
|
|
name: freebsd-cost-optimization
|
|
description: How to use Colibri's fast/smart/max cost modes to minimize DeepSeek API spending on FreeBSD deployments
|
|
---
|
|
|
|
# FreeBSD Cost Optimization
|
|
|
|
## Cost modes
|
|
|
|
Colibri supports three cost modes that control context window usage:
|
|
|
|
- **fast** — tighter context budget, compaction triggers earlier. Use for
|
|
high-volume, low-importance tasks (log scanning, status checks).
|
|
- **smart** (default) — balanced budget. Most agent sessions should run here.
|
|
- **max** — preserves full context, compaction only when absolutely necessary.
|
|
Use for complex multi-step tasks where losing history would cause rework.
|
|
|
|
## When to escalate
|
|
|
|
The daemon auto-escalates fast→smart→max when compaction alone cannot keep the
|
|
session within budget. This is intent, not a policy engine — the thresholds are
|
|
fixed in `colibri-daemon/src/cost.rs`:
|
|
|
|
- fast: budget 4K tokens, compact at 80%
|
|
- smart: budget 16K tokens, compact at 75%
|
|
- max: budget 64K tokens, compact at 90%
|
|
|
|
## Operator tips
|
|
|
|
- Set `COLIBRI_COST_MODE=fast` in `/etc/rc.conf` for disposable worker agents.
|
|
- Use `colibri set-cost-mode smart` via the CLI for interactive sessions.
|
|
- Monitor cache-hit rate via `colibri status` — if it drops below 50%,
|
|
consider escalating to the next mode.
|
|
- On FreeBSD live USB, the default is `smart` to balance cost and capability.
|