Introduction

app.csvibes is a tool that allows you to apply multiple artificial intelligence
models simultaneously to a dataset, generating files, text, and images using multiple AI models working together.
This enables you to obtain advanced artificial intelligence with many use cases.
This page explains the various functions. To load a dataset, you need to upload a file in .csv, .json, or .xml format,
which will be displayed in a table.
Subsequently, you can manipulate and complete the dataset or use it as a basis to create multiple text files, create
images, or create files containing text and images using the template function. Csvibes File Upload

Upload a File

To upload a file containing a dataset, you need to first log in and then click on the Select & Upload File” button. The system accepts the
formats .csv, .json, and .xml.
Afterward, select the file to upload and click OK. The file will be displayed in the table.

Write and Execute a Prompt

Once you’ve loaded the dataset into the table, you will find a button ” >_ “ above each column,
as shown in the following image.
Write prompts on Csvibes

By clicking on the ” >_ ” button, a window like the one in the following image will open. Here,
you can write the prompt and choose the model to use for that column.

In Csvibes, you can always insert variables taken from the columns for each row. To insert variables, you use the
notation {COLUMN_LETTER}.
An example of a prompt could be: “write a description of {A} in {B} language” where columns A and B contain lists of
names and languages; Csvibes will execute the prompts using data from columns A and B for each row.
Variables can be used in prompts for both text and images and inside the template file or the merge
function,
as explained in detail in the following sections.

Upload file Csvibes

In the window, there are 3 tabs:

1. Prompt – In the “Prompt” tab, you can write a command to generate text (executes one of the LLM
models).
2. Image – In the “Image” tab, you can write a command to generate images (executes one of the
Stable diffusion models).
3. Template – In the “Template” tab, you can upload a template file to generate a file.

In the following sections, the 3 tabs for text generation (Prompt), image generation (Images), and file generation
(Template) will be explained individually.

Texts – Images – Templates – Merge

The 3 tabs, prompts, images, and templates, can be applied to all columns, and different functions can be performed
on different columns. Csvibes will execute the functions on all columns row by row.

Texts

After clicking the ” >_ “ button, a window opens, and by selecting the
Prompt tab, you can apply Large Language Model (LLM) text generation models to all rows in the
column.
Write prompts on Csvibes

In the “Prompt” tab window, you will find the following fields:

1. Title – The title is the title of the column and corresponds to the column key that will be
used as the key when exporting the dataset to .csv, .json, or .xml format.

2. Prompt – In the prompt field, you can write a natural language command that will be executed
by the Large Language Model artificial intelligence model for text generation. Write in a simple and colloquial
manner what you want the AI to execute row by row.

In this field, you can use the values from other columns as variables using the format
{COLUMN_LETTER}.

An example of a prompt for LLM in natural language could be: “write text about {A} in relation
to {B}.” Csvibes will execute the command for each row in the table, replacing
{A} and {B} with the values from the corresponding columns for each row and
writing the responses in each row of the column.

Images

After clicking the ” >_ “ button, a window opens, and by selecting the Images
tab, you can apply image generation models using Stable Diffusion models to all rows in the column.

Create Images

Create images on Csvibes

In the “Image” tab window, you will find the following fields:

1. Title – The title is the title of the column and corresponds to the column key that will be
used as the key when exporting the dataset to .csv, .json, or .xml format.

2. Prompt – In the prompt field, you can write a natural language command that will be executed
by the artificial intelligence model for image generation. Describe in a simple manner the image description you
desire.

In this field, you can insert values from other columns using the format {COLUMN_LETTER}.
An example of an image prompt could be: “Create an image of {C}.” Csvibes will execute the
command for each row in the table, replacing {COLUMN_LETTER} with the values from the
corresponding columns for each row and inserting the image into each row of the column.

Templates

After clicking the ” >_ “ button, a window opens, and by selecting the
Templates tab, you can generate a file for each row in the column.

Create Files

Create templates on Csvibes

In the “Template” tab window, you will find the following fields:

1. Title – The title is the title of the column and corresponds to the column key that will be used
as the key when exporting the dataset to .csv, .json, or .xml format.

2. Template file – In the Template file field, you can upload a template file in text formats: txt,
html, py, js, php, by including the values from other columns as variables within the file using the format
{COLUMN_LETTER}.

Save Generated Files

By clicking the Save button in Csvibes, it will create a file for each row of the table, replacing
{COLUMN_LETTER} with the texts or images (URLs) from the corresponding columns for each row within
the files.
Create templates on Csvibes

Clicking on the column title “Template” on the ” ” icon will open a new window
with the following fields:

1. File Name – Choose the column to use for assigning names to the files

2. Extension – In the Extension field, you can choose the file extension from the available
options: .pdf – .docx (MS Word) – .txt – .css (Style) – .php – .asp – .html – .py – .js (JavaScript) – .class –
.java (Java) – .vb – .c.

Create templates on Csvibes

Within the Template column, two buttons will appear on each row:
Tpl Preview ” where you can preview the generated file
” where you can save the individual file locally.

The generated file will contain all the responses generated by various models corresponding to the columns used as
variables, which, when combined with the text contained in the template, will create the files row by row.

Template generation is completely free and does not consume any credits.

If you export the dataset to .csv, .json, or .xml, the Template column will contain the names of the files complete
with extensions.

Prices

Storage and credits

1 credit = 1 cell

200 credits = Free for every new user.

5000 credits: 5 USD

Storage beyond 1 month: 9 USD/month 3GB

Large language models

GPT 3.5 Turbo: 1 credit per created cell.
GPT 4: 3 credits per created cell.
Cloude 2: 2 credits per created cell.
Cohere Command: 2 credits per created cell.
Meta Llama 2: 0.5 credits per created cell.
Other transformers: FREE
File creation: FREE

Stable diffusion

From 2: 12 credits per created image
Midjourney 4.0: 5 credits per created image
F22: 1 credit per image
XSarchitectural-InteriorDesign: 1 credit per image