How to Manage AI Usage Costs in Folian

Folian is designed around a bring-your-own AI keys model.

That means you connect your own supported AI provider, such as OpenAI or Claude where available, and manage AI usage through that provider.

This gives writers more control, but it also means usage should be understood before heavy writing sessions.

How AI usage is usually charged

AI providers usually charge based on usage.

The final cost can depend on:

• The model used.
• The amount of text sent to the model.
• The amount of text generated.
• How often AI is used.
• Whether the task is short or context-heavy.

A quick brainstorming prompt is usually lighter than asking AI to review a long chapter with style rules and story bible context.

Why writing tasks vary in cost

Different writing tasks need different amounts of context.

A simple idea request may only need a short prompt.

A chapter revision task may need the chapter text, style guidance, character notes and canon context.

That means two AI requests can look similar to the writer but have very different usage behind the scenes.

How to stay in control

• Start with small tests.
• Check your AI provider dashboard regularly.
• Use lighter models for simple tasks where possible.
• Use stronger models only when the task needs them.
• Set usage limits where your provider supports them.
• Review billing before long sessions.
• Avoid sending large amounts of context when a small request will do.

How Folian helps

Folian is being built to use AI as controlled support rather than constant generation.

The writer should be able to choose when AI is useful, review outputs before applying them and keep AI inside the structure of the project.

That makes usage more intentional.

When to use AI lightly

Use lighter AI support for:

• Brainstorming.
• Small rewrites.
• Idea comparison.
• Chapter questions.
• Short critique.
• Planning notes.

When stronger AI may be useful

Stronger AI support may be useful for:

• Long chapter analysis.
• Style review.
• Continuity checks.
• Complex revision.
• Story bible summarisation.
• Cross-project reasoning.

The writer should decide whether the task is worth the usage.

Need help connecting your key? Read Bring Your Own AI Keys.