Rework model tiers: Haiku default with internal routing to Ollama/Sonnet

This commit is contained in:
Jarvis
2026-02-05 10:35:59 -07:00
parent 10c678ee39
commit 30658a9cff

View File

@@ -11,34 +11,39 @@
**Monthly Budget:** **Monthly Budget:**
- **$20 total** - **$20 total**
- **Warn at 75%** ($15 spent) - **Warn at 75%** ($15 spent)
- **Ollama tasks are free** — prioritize for routine work
--- ---
## MODEL SELECTION ## MODEL SELECTION
### Tier 1: Ollama (Free Local) ### System Default: Haiku
Use for:
Haiku is your primary model. It decides routing internally:
**Haiku routes to Ollama when:**
- File checking and organization - File checking and organization
- Heartbeat tasks (status checks, log review)
- Simple templating/formatting - Simple templating/formatting
- Log review and cleanup
- Non-critical analysis - Non-critical analysis
- Routine status checks
**Advantage:** Free, instant, zero API cost **Haiku handles directly when:**
- Most tasks fit within Haiku's capability
- Reasoning is needed but not deeply complex
- Code review (non-production)
- Documentation and writing
### Tier 2: Default - Haiku **Haiku escalates to Sonnet when:**
Switch to **Sonnet** ONLY when:
- Architecture decisions - Architecture decisions
- Production-like code review - Production-like code review
- Security analysis - Security analysis
- Complex debugging/reasoning - Complex debugging/reasoning
- Strategic multi-project decisions - Strategic multi-project decisions
### Heartbeat: Ollama Only
Heartbeats ALWAYS use Ollama. No escalation. If Ollama fails, the heartbeat fails.
### Decision Rule ### Decision Rule
- **Ollama first** for routine/repetitive work Let Haiku decide. It's smart enough to route to Ollama when appropriate and escalate to Sonnet when needed. You only override when you know you need Sonnet upfront.
- **Haiku second** for most other tasks
- **Sonnet last** for genuinely complex thinking
**When in doubt:** Try Ollama first, escalate if needed.