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

# Fast and detailed planning

> What the MESO Think toggle changes, and when the extra wait is worth it.

Every unit is generated one of two ways. The toggle is called **MESO Think** and
sits on the create-unit screen, just before you generate.

## The two modes

| | Standard | MESO Think |
| - | - | - |
| **On by default** | Yes | No |
| **How it works** | A single pass | Separate agents draft, critique and refine the plan |
| **Depth** | Solid working draft | More depth |
| **Speed** | Faster | **Slower — expect a wait** |
| **Cost** | Standard | **Uses extra credits** |

That is the whole of the difference. Both modes plan against the same
specification and the same outcomes you selected; Think spends more effort on
the draft rather than covering anything different.

## When to use Think

Worth it when:

* the unit matters more than usual — a first run of a new topic, or something
  you will reuse across years
* you want the reasoning behind the sequence to be more considered
* you are not waiting on it to teach this afternoon

Not worth it when:

* you are drafting quickly and intend to rewrite heavily anyway
* you just want a sequence to react to
* you are generating several units in a sitting

> [!TIP]
>
> Standard is the sensible default, which is why it is the default. Reach for
> Think deliberately rather than habitually — it costs both time and credits.

## Either way, you still review it

> [!WARNING]
>
> Detailed does not mean verified. A Think-generated plan can be confidently wrong
> in exactly the same ways as a standard one — dates, quotations, formulae and
> worked answers all still need checking, and cited outcome references still need
> verifying against the specification.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.