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

# Tokens

> How Mythex tokens work — grants, top-ups, markup, and what happens when you run out.

# Tokens

Tokens power the **coding agent**. Every agent run consumes tokens based on model usage.

## Grants

| Source               | Rough size                                           |
| -------------------- | ---------------------------------------------------- |
| Free monthly credits | \$2 → about **1,000,000** tokens                     |
| Pro monthly included | \$15 → about **7,500,000** tokens                    |
| Top-ups              | \$5 / \$10 / \$25 one-time packs (do **not** expire) |

Conversion used in product: **500,000 tokens ≈ \$1** (1¢ = 5,000 tokens).

## How charging works

Mythex charges based on underlying model cost with a product **markup** (about **1.6×** on raw model USD). Small usages still bill at least **1 token** when cost is positive.

## Running out

When the balance hits zero:

* The agent **pauses**
* Projects, files, and published apps **remain**
* Top up or wait for the next monthly grant (Free or Pro)

## Tips to spend efficiently

* Prefer precise prompts over “rebuild everything”
* Use Plan mode when exploring large changes
* Fix Preview errors with targeted asks

## Related

* [Plans](/billing/plans)
* [Manage billing](/billing/manage)
* [Effective prompts](/prompting/effective-prompts)