Skip to main content

How Word Usage is Calculated for Chats

Written by Dustin W. Stout
Updated today

At Magai, we offer various subscription plans based on the number of words you can generate per month using our AI models. Understanding how word usage is calculated is key to managing your plan and maximizing your productivity. Here’s a detailed guide to help you navigate this.

Word Usage Calculation

Your word usage is calculated based on the number of words generated by the AI model, multiplied by a specific factor known as the AI model multiplier. This multiplier varies depending on the AI model you are using.

Formula:

[number of words generated] * [AI model multiplier] = [words used]

AI Model Multipliers

Each AI model available in Magai has a distinct multiplier. Here are examples of some popular models:

GPT-5.4: Multiplier of 1.2

Example: If you generate 100 words using GPT-5.4, it will consume 100 words from your monthly quota. [100 words] * [1.2] = 120

Gemini 3 Pro: Multiplier of 1

Example: If you generate 100 words using Gemini Pro, it will consume 10 words from your monthly quota. [100 words] * [1] = 100

Below is a more detailed look at how different multipliers affect your word usage:

Benefits of Different Multipliers

Using models with lower multipliers allows you to generate more content within your monthly word quota. Understanding the multipliers attached to each AI model will help you choose the right model for your needs and manage your plan more effectively.

Full List of Models and Multipliers

Name

Type

Multiplier

Auto

LLM

1x

Claude Haiku 4.5

LLM

0.4x

Claude Opus 4.6

LLM

2x

Claude Opus 4.7

LLM

2x

Claude Sonnet 4

LLM

2x

Claude Sonnet 4.6

LLM

1.2x

DeepSeek V3

LLM

0.1x

DeepSeek V3.2

LLM

0.04x

Flux

Image

1x

Flux Kontext Max

Image

1.3x

Flux 2 Standard

Image

1x

Flux 2 Pro

Image

1x

Flux 2 Max

Image

1x

Flux 2 Klein

Image

1x

Flux 2 Klein Base

Image

1x

Gemini 2.5 Flash

LLM

0.2x

Gemini 2.5 Pro

LLM

0.8x

Gemini 3 Pro

LLM

1x

Gemini 3.1 Flash Lite

LLM

0.3x

Gemini 3.1 Pro

LLM

2x

GLM 5

LLM

0.22x

GLM 5 Turbo

LLM

0.28x

GPT Image

Image

5x

GPT Image v1

Image

4.2x

GPT Image v1.5

Image

3.3x

GPT Image v2

Image

3.3x

GPT OSS 120B

LLM

0.02x

GPT-5 Image

LLM

1.3x

GPT-5.2 Fast

LLM

2x

GPT-5.4

LLM

1.2x

GPT-5.4 Mini

LLM

0.35x

GPT-5.4 Nano

LLM

0.1x

GPT-5.4 Pro

LLM

14x

Grok 3

LLM

2x

Grok 3 Mini

LLM

0.1x

Grok 4

LLM

1.2x

Grok 4.1 Fast

LLM

0.05x

Grok 4.20

LLM

0.53x

Grok 4.20 Multi-Agent

LLM

0.53x

Hailuo Standard

Video

5x

Hailuo Pro

Video

8x

Ideogram

Image

2x

Kimi K2.5

LLM

0.23x

Kling v3 Image

Video

1x

Kling v3

Video

4x

Kling v3 Pro

Video

5x

Kling v3 Edit Video

Video

4x

Kling v3 Pro Edit Video

Video

5x

Kling Img Ref

Video

4x

Kling v3 Pro Img Ref

Video

5x

Kling Vid Ref

Video

4x

Kling Pro Vid Ref

Video

5x

Leonardo Motion

Image

2x

Leonardo.ai

Image

2x

Llama 4 Maverick

LLM

0.05x

Llama 4 Scout

LLM

0.03x

Luma

Video

9x

MiMo V2 Omni

LLM

0.16x

MiMo V2 Pro

LLM

0.27x

MiniMax

Video

9x

MiniMax M2.5

LLM

0.1x

MiniMax M2.7

LLM

0.1x

Mistral Large 3

LLM

0.13x

Mistral Pixtral

LLM

0.6x

Mistral Small 4

LLM

0.05x

Nano Banana

Image

1x

Nano Banana v2

Image

1.3x

Nano Banana v1 Pro

Image

2.5x

Nano Banana v1 Standard

Image

0.7x

Nemotron 3 Nano

LLM

0.02x

Nova 2 Lite

LLM

0.19x

Nova Pro

LLM

0.3x

o4 Mini

LLM

0.4x

o4 Mini Deep Research

LLM

0.7x

Perplexity Deep Research

LLM

0.7x

Perplexity Sonar

LLM

0.2x

Perplexity Sonar Pro

LLM

1.2x

Perplexity Sonar Pro Search

LLM

1.2x

Recraft

Image

1x

Reve

Image

1x

Runway v4

Image

10x

Runway V4 Aleph

Image

10x

Runway v4.5

Image

10x

Runway v4 Image

Image

1x

Runway Image Turbo

Image

1x

Seedance v2

Video

10x

Seedance v2 Fast

Video

10x

Seedance v2 Reference

Video

10x

Seedance v2 Reference Fast

Video

10x

Seedance Lite

Video

3x

Seedance Pro

Video

10x

Seedream v5

Image

1x

Seedream v4

Image

0.5x

Sora 2 Standard

Video

13.3x

Sora 2 Pro

Video

40x

Veo v3

Video

53.3x

Veo v3 Fast

Video

20x

Veo v3.1

Video

26.7x

Veo v3.1 Fast

Video

13.3x

Voice GPT

LLM

1x

Managing Your Plan

Here’s how you can manage your plan to maximize your monthly word quota:

Choose the right AI model: Depending on your needs, select a model with a suitable multiplier. For producing large volumes of text, models with lower multipliers like Gemini Pro are ideal.

Monitor your word usage: Keep track of the number of words generated by each model to ensure you stay within your monthly limit.

Upgrade or adjust your plan: If you find that you consistently exceed your word quota, consider upgrading your plan for a higher word limit or adjusting your usage patterns.

Conclusion

Understanding how word usage is calculated and the multipliers assigned to different AI models is crucial for managing your subscription plan effectively. By choosing the right model and keeping track of your usage, you can make the most out of your Magai subscription.

If you have any questions or need further assistance, feel free to contact our support team. Happy creating with Magai!

Did this answer your question?