Complexity Tier
Published Updated
What is Complexity Tier?
A complexity tier is the label on an Orbit prompt (Low, Medium, High or Ludicrous) that decides which coding agent and model run it. You map each tier to an agent and model once, in settings.
A complexity classification (low / medium / high) on each prompt that determines which model and agent runs it. Configurable in Settings.
Every queued prompt declares a `complexity:` tag (low, medium, high). Settings → Execution Tiers maps each tier to a specific model and agent. Low routes to the cheapest configured option; high routes to the most capable. The user-defined config means routing is deterministic, not subject to model interpretation.
How it relates to Orbit
A prompt’s complexity tag selects the agent and model tier configured in Orbit settings. See Complexity Tiers & Routing for details.
When you would use it
You configure Execution Tiers once during setup and adjust as your workflow evolves. The tier on each prompt determines what runs it.
Common operations
- Configuring which model runs each tier
- Routing routine prompts to cheap models
- Routing complex prompts to capable models
Frequently asked questions
What is the difference between complexity: and model:?
complexity: routes through your tier settings; model: names a model directly. They are mutually exclusive, and if both are present complexity wins. See Complexity Tiers & Routing.
Can a tier use a local model?
Yes. Any tier can route to Ollama, LM Studio, DeepSeek, MiniMax or a custom OpenAI-compatible base URL.
Related terms
Dashed terms are named in the source but do not have a page here yet.
Where this is taught
No learning path uses this term yet. Browse the Learning Atlas for guided sequences through related ideas.