Canonical page: https://mixerai.org/en/knowledge/text/gpt-text/

[Mixer AI](https://mixerai.org/) › [Knowledge Base](https://mixerai.org/en/knowledge/) › [Text](https://mixerai.org/en/knowledge/text/) › GPT

# GPT

Turn a brief, notes and images into a draft, plan or clear set of decisions.

[Launch GPT in Mixer AI](https://mixerai.org/?dl=kb_gpt-text#/canvas)

## When to choose it

The GPT family in Canvas fits the steps between creative tasks: unpack a brief, prepare a script, compare references or rewrite for an audience. Choose a version around complexity and budget rather than automatically selecting the most expensive. Clear inputs and an explicit output format make the answer easier to verify.

- Discuss text and images within the same task.

- Compare different versions on an identical brief, including their cost.

## Modes and settings

- GPT 6 Astra, GPT 5.6 Sol, Terra and Luna, selected before a Canvas request.

- Connect up to 5 text blocks and 10 images. This tool produces text, not a new image.

## Price in coins

| Version | Input, coins / 1M tokens | Output, coins / 1M tokens
| GPT 6 Astra | 361.2 | 1806
| GPT 5.6 Sol | 180.6 | 1083.6
| GPT 5.6 Terra | 72.24 | 433.44
| GPT 5.6 Luna | ≈ 7.22 | ≈ 43.34

Rates are in coins per million tokens, not per reply. A token is a piece of text, not necessarily a word. Input and output have separate rates. Canvas reserves an estimate and settles from actual usage; images also affect usage.

≈ marks a fractional rate rounded to two decimal places for display; billing uses full precision.

## What you can make

- Turn scattered comments into requirements for a video.

- Prepare a script, prompt or edit while preserving the supplied facts.

## How to direct the result

- Start with the deliverable, audience and required format.

- Separate facts, style samples and preferences. Ask it not to invent missing information.

### Example: comments into an editing plan
Copy

```text
Review the connected brief and client comments. Return three sections: required edits, contradictions and questions before editing. Do not invent deadlines or a budget. For each edit, identify the source requirement. Preserve numbers and names. Stay under 2500 characters.
```

## What to keep in mind

- Canvas saves answers up to 3400 characters. Work on larger material in sections.

- The pre-request amount is an estimate; final billing uses actual usage. Answers can contain errors.

## FAQ

Which version should I use for simple writing?

For a short edit or format conversion, start with a lower-priced version in the table. Compare Astra or Sol on tasks with several constraints.

Can I analyze photos?

Yes. Connect the relevant images and specify whether to compare composition, visible details or fit with the brief.

Does this GPT tool generate pictures?

No. It returns text. Use the separate GPT Image page to create images.

Can I write a long article in one request?

Canvas limits the saved answer length. Plan a larger article first, then work section by section.

Will it open a link in the prompt automatically?

Do not rely on that. Add the required text or images directly to the task.

Why does a larger context cost more?

Input and output volume matter, and extra images add work. Connect only the material you need.

What changes the price?

It depends on the version, the text and images supplied, and the response length. Canvas reserves an estimate first, then bills actual usage. The rates in the table help compare versions; they are not a fixed price for every reply.

Do I need a separate model subscription?

No. In Mixer AI, you top up one coin balance and use it for the tasks you need. You do not need a separate subscription to this model.

[Launch GPT in Mixer AI](https://mixerai.org/?dl=kb_gpt-text#/canvas)