> ## Documentation Index
> Fetch the complete documentation index at: https://docs.odella.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Model & Intelligence

> Choose the model, reasoning level, and infrastructure size for a workforce employee.

**Model & Intelligence** controls the worker's core thinking and runtime capacity.

## LLM Model

The default language model used by the worker.

Models vary by provider, speed, cost, and reasoning quality. The selector displays a cost indicator such as `$`, `$$`, `$$$`, or `$$$$`.

Use lower-cost models for routine work and stronger models for complex analysis, drafting, or planning.

## Reasoning Effort

Controls how much reasoning the model should apply before responding or acting.

| Value      | Best for                                         |
| ---------- | ------------------------------------------------ |
| **Off**    | Fast, simple responses.                          |
| **Low**    | Straightforward work with light reasoning.       |
| **Medium** | Balanced default for most workers.               |
| **High**   | Complex analysis, planning, or careful drafting. |

Higher reasoning effort can increase token usage and processing time.

## Infrastructure Size

Controls the worker's instance size and concurrency budget.

| Size                 | Best for                                                                                            |
| -------------------- | --------------------------------------------------------------------------------------------------- |
| **Small Instance**   | Personal workflows or a narrowly shared assistant. Typically 1–3 people and 1 live conversation.    |
| **Medium Instance**  | A shared team assistant used throughout the workday. Typically 4–8 people and 2 live conversations. |
| **Large Instance**   | Broad team coverage or higher-volume shared use. Typically 10–20 people and 4 live conversations.   |
| **X-Large Instance** | Heavy enterprise use. 20+ people and 8 live conversations. Contact support for dedicated scaling.   |

Infrastructure size contributes to the effective hourly rate shown in Billing.

<Tip>
  Start with the smallest instance that fits the expected usage, then increase size when the worker is shared more broadly or needs more concurrent conversations.
</Tip>