> For the complete documentation index, see [llms.txt](https://helpcenter.paperguide.ai/llms.txt).
> Use it to discover all available pages before exploring further.

# AI Credits

Understand how AI credits work in Paperguide, how many credits you receive each month by plan, and how to estimate usage for key AI features.

## How AI credits work

Paperguide agents and AI features consume credits whenever they run tasks such as literature reviews, deep research reports, or data extraction. Each request uses a certain number of credits based on how much AI processing is required.

Credits are tied to your Paperguide plan and refresh every month on your billing date.

### Features that consume AI credits

The following features use AI credits each time they run:

- Systematic Review
- Deep Research Report
- Extract Data
- Research Agent Chats (Research agent, Literature Review Agent, AI Chats in Writer and Reference Manager)
- AI Writer - Generate Document
- AI Writer – AI Edit
- MCP Server — Search papers, search inside your PDFs, and read papers from your AI assistant

<Callout kind="info" collapsed="false">
  The MCP server uses the same AI credit pool as all other Paperguide features. Searching all papers costs 10 credits, fetching one paper's details 2, searching inside your PDFs 10, and reading a paper's full text 10 per part. Finding, browsing, and organising your saved references is free. Failed calls are not charged. See [MCP credits](/mcp/credits) for the full table.
</Callout>

## Estimating AI credit usage

Use the examples below to understand what you typically get each month from your AI credits on each plan.

Here's a quick summary of estimated monthly usage by plan:

| Feature           | Scenario              | Plus (12,500 credits) | Pro (50,000 credits) | Max (150,000 credits) |
| ----------------- | --------------------- | --------------------- | -------------------- | --------------------- |
| Systematic Review | 1,000 papers screened | Up to 2 reviews       | Up to 10 reviews     | Up to 30 reviews      |
| Systematic Review | 5,000 papers screened | Not feasible          | Up to 2 reviews      | Up to 6–7 reviews     |
| Literature Review | Per month             | \~15–17 reviews       | \~60–70 reviews      | \~180–200 reviews     |
| Extract Data      | 10 fields per paper   | 300–400 papers        | 1,200–1,600 papers   | 3,600–4,800 papers    |

<Callout kind="info" collapsed="false" color="#f59e0b">
  The numbers below are approximations. Actual usage may vary depending on task complexity, content length, references included and the specific AI features you use.
</Callout>

### Systematic review

A systematic review starting with **1,000 papers screened:**

With **12,500 credits/month (Plus)**, you can run **up to 2** systematic reviews of this scale each month.

With **50,000 credits/month (Pro)**, you can run **up to 10** systematic reviews of this scale each month.

With **150,000 credits/month (Max)**, you can run **up to 30** systematic reviews of this scale each month.

### Systematic review

A systematic review starting with **5,000 papers screened:**

With **12,500 credits/month (Plis)**, a full 5,000-paper systematic review is not feasible.

With **50,000 credits/month (Pro)**, you can run **up to 2** systematic reviews of this scale each month.

With **150,000 credits/month (Max)**, you can run **up to 6-7** systematic reviews of this scale each month.

### Literature review

With **12,500 credits/month (Plus)**, you can get around **15-17 literature reviews** spread across your projects each month.

With **50,000 credits/month (Pro)**, you can get around **60-70 literature reviews** for larger or more frequent research workflows.

With **150,000 credits/month (Max)**, you can get around **180-200 literature reviews** for larger or more frequent research workflows.

### Extract data

A extract data table with approximately **10 fields/columns per paper extracted:**

With **12,500 credits/month (Plus)**, you can get around **extract 300 - 400 papers&#x20;**&#x77;ith 10 fields per paper across your workbooks each month.

With **50,000 credits/month (Pro)**, you can get around **extract 1,200 - 1,600 papers&#x20;**&#x77;ith 10 fields per paper across your workbooks each month.

With **150,000 credits/month (Max)**, you can get around **extract 3,600 - 4,800 papers** with 10 fields per paper across your workbooks each month.