AxonSheet AI Assistant

Welcome to AxonSheet

i Your intelligent assistant for Excel. Automate tasks, gain insights, and boost your productivity.

🤖

AxonSheet AI Assistant Ready

{/* Content will be populated by JS */}

Create a New Table

Describe the data you need and let AI build it.

Add Rows to Table

Expand your tables with more data, guided by AI.

Add Columns to Table

Let AI analyze your table and suggest new columns.

Fill Empty Cells

Intelligently complete empty cells in your tables.

Edit Selected Cells

Rewrite cell content based on your instructions.

Audit & Comment Cells

Have AI review cells and add comments with feedback.

Process Data Rows

Analyze, categorize, or transform data row by row.

i Describe the table structure and content you need (e.g., '5 largest planets by diameter'), and the AI will generate the headers and data rows.

1 Define Your Input

Text Input and Instructions

Allow Formula Generation

Use Excel Selection as Context

No range selected.

Use Case

Invoice Contract CV

Use Document Content

Drop file here or click to upload

Supported: TXT, DOCX, PDF, CSV, MD

📄

Conversion Method:

Using default OCR model from Settings. Change the model in the Settings section.

Converted Document Snippet:

Enable Deep Research

Use table description as research input

Use document content as research input

Deep Research Results:

2 Plan Your Output Table(s)

i Select an existing table. Describe new rows or provide context via documents/research. AI generates new rows matching your table structure.

1 Select Your Range

First select the table you would like to add new rows to.

No range selected.

2 Define Your Input

Add Rows Settings

Sample Rows from Existing Table for AI Context

Describe New Rows

Use Document Context

Drop file here or click to upload

Supported: TXT, DOCX, PDF, CSV, MD

📄

Conversion Method:

Using default OCR model from Settings.

Converted Document Snippet:

Enable Deep Research

Use selected table data as research context

Use row description as research input

Use processed document as research input

Deep Research Results:

Research & Document Interaction

Run in Parallel (Default) Sequential (Document Informs Research)

i Select your table data (including headers). The AI analyzes the content and suggests new, related columns. Choose 'Automatic' analysis for AI-driven column suggestions or 'Define Every Column' to define specific new columns.

1 Select Your Range

First select the table range that you would to add new columns to.

No range selected.

2 Define Your Input

Text Input and Instructions

Allow Formula Generation

Use Document Content

Drop file here or click to upload

Supported: TXT, DOCX, PDF, CSV, MD

📄

Conversion Method:

Using default OCR model from Settings. Change the model in the Settings section.

Converted Document Snippet:

Enable Deep Research

Use selected table data as research context

Use document content as research input

Deep Research Results:

3 Plan Your Output New Columns

i Select a range or table (including headers). Click 'Read Data' to load the selection, then choose the columns you want the AI to fill with inferred values for empty cells.

1 Select Your Range

First select the table range with blank cells you would like to fill.

No range selected.

2 Define Your Input

Fill Empty Cells Settings

Header Location:

Always fill all empty cells (skip column selection)

i If checked, Cell Filler will automatically target all columns with empty cells in your selection and proceed without asking for column confirmation.

Optional Instructions

Allow Formula Generation

Use Document Context

Drop file here or click to upload

Supported: All document types

📄

Conversion Method:

Using default OCR model from Settings.

Document Instructions (Optional)

Converted Document Snippet:

Enable Deep Research

Use main "Optional Instructions" as research context

Use selected table data as research context

Deep Research Output:

i Select multiple ranges, choose a goal, and let AI intelligently merge, stack, or consolidate your data into a single, clean table.

1 Select Your Ranges

Use Ctrl+Click to select the tables or ranges you want to merge.

No range selected.

2 Choose Your Consolidation Goal

Merge and Create Master Table Standardise Anonymize & Create Shareable Copy

3 Refine with Instructions & Documents

Additional Instructions (Optional)

Allow Formula Generation

Use Document Content

Drop file here or click to upload

Supported: TXT, DOCX, PDF, CSV, MD

📄

Conversion Method:

Using default OCR model from Settings. Change the model in the Settings section.

Converted Document Snippet:

Enable Deep Research

Use selected table data as research context

Use document content as research input

Deep Research Results:

4 Plan Your Output (Optional)

i Select a table or range. Click 'Edit Selected Cells'. Provide instructions on how the AI should modify the data (e.g., 'Correct typos in column A'). Preview and apply changes.

1 Select Your Range

First select the table range you would like to edit.

No range selected.

2 Define Your Input

Edit Selected Cells Settings

Header Location:

Optional Instructions

Allow Formula Generation

Use Document Context

Drop file here or click to upload

Supported: All document types

📄

Conversion Method:

Using default OCR model from Settings.

Document Instructions (Optional)

Converted Document Snippet:

Enable Deep Research

Use main "Optional Instructions" as research context

Use selected table data as research context

Deep Research Output:

i Select a table or range. Click 'Read Data for Reviewing'. Provide instructions on what the AI should review. The AI will highlight cells and add comments with feedback.

1 Select Your Range

First select the table range you would like to review.

No range selected.

2 Define Your Input

Audit & Comment Cells Settings

Header Location:

Clear existing colors & comments in target columns before applying new review

i If checked, any existing cell background colors and comments in the columns you select for review will be removed before the new feedback is applied.

Optional Instructions

Allow Review of Formulas

Use Document Context

Drop file here or click to upload

Supported: All document types

📄

Conversion Method:

Using default OCR model from Settings.

Document Instructions (Optional)

Converted Document Snippet:

Enable Deep Research

Use main "Optional Instructions" as research context

Use selected table data as research context

Deep Research Output:

i For every row in your selection, the tool constructs a prompt by combining your 'Before Text', the text from the selected cells, and your 'After Text'. This complete prompt is then sent in a separate request to the AI. Results are placed in your chosen output column. Ideal for tasks applied row-by-row.

Rows Processor Settings

1 Select Your Range

First select the data rows you would like to process.

No range selected.

2 Define Your Input

Prompt Prefix

0 characters

Prompt Suffix

0 characters

Header Position is at

Top of Selection Range First Row of Sheet No Header

Enable Deep Research (per row)

Uses the global Research Engine in Settings. For each row, the add-in runs deep research, then appends findings to the AI prompt.

Allow Formula Generation

3 Plan Output

Define Output Fields

i Define each output field. This will be used as the column header and as an instruction for the AI. Leave empty for single output.

Skip Rows if Output Exists

Maximum Concurrent Requests

Output Column

4 Review Input to AI

{/* Preview content will go here */}

Sample 0 of 0

Legend: Prefix | Cell Header: Cell Value | Suffix | Multi-Output Instruction

i Process multiple documents to extract information and add it to your table.

1 Select Your Range

First select the table you would like to add new rows to.

No range selected.

2 Define Your Input

Add Rows

Add Columns

Fill Empty Cells

Edit Selected Cells

Multi Document Processor Settings

Sample Rows from Existing Table for AI Context

Custom instruction

Allow Formula Generation

Drop file here or click to upload

Supported: TXT, DOCX, PDF, CSV, MD

📄

Conversion Method:

Using default OCR model from Settings.

Converted Document Snippet:

Import & Convert

i Select a text, Word, or PDF file. Choose a conversion method to extract tabular data. Decide whether to create a new table or append to the currently selected one.

Import Data Settings

Drop file here or click to upload

Supported: TXT, DOCX, PDF

📄

Conversion Method:

Vision Model

🔍

Target Mode: Create New Table Append to Selected Table

Optional Instructions:

{/* Hidden by default, shown if append mode */}

Sample Rows for Append Context

{/* Renamed ID from tgResearchDocInteractionOptions */} Research & Document Interaction (if applicable): Run in Parallel Sequential (Document Informs Research)

i Paste your JSON data (usually an array of objects) into the text area below. Click 'Convert to Table', and the data will be written to the active sheet.

JSON to Table Settings

JSON Data:

i Select a table range in Excel (including headers) and click 'Generate Schema'. You can add optional notes to guide the AI.

JSON Schema Settings

Optional Instructions for AI:

i Paste Markdown text containing one or more tables. The tool will extract each table and create corresponding tables in Excel.

Markdown to Table Settings

Paste Markdown Text:

i Paste Markdown text. Choose to 'Extract Tables Only' to get standard Excel tables, or 'Convert Entire Document' to render the full Markdown structure into Excel cells.

Markdown Import Settings

Paste Markdown Text:

Enable entire document conversion

If unchecked, only tables will be extracted. If checked, the entire document structure will be converted.

i Paste Mermaid diagram code to generate a visual diagram in Excel.

Mermaid Settings

Mermaid Diagram Code

i Paste Cytoscape JSON to generate a visual diagram in Excel. The renderer will create shapes and connectors based on the 'excel' block in your data.

Cytoscape Settings

Cytoscape JSON Input:

Add-in Settings

i Configure your connection to AI services like OpenRouter, Ollama, or OpenAI. Select default models for various tasks.

AI Provider

OpenRouter Configuration

Default Task Model

🔍

API Key

Enable Real-Time Web Search

i Integrates live web search results into AI responses. Additional cost: $4 per 1,000 requests.

Ollama Local AI Configuration

Checking for Ollama server...

Select Model

Requires Ollama server running locally. Setup Help

OpenAI Configuration

Enter API Key to validate.

API Key

Your key is stored locally using Office settings.

Select Model

Choose the OpenAI model to use.

Custom AI Service Configuration

Enter API Key, Base URL, and Model ID.

API Key (Optional)

Your key is stored locally.

Base URL

The endpoint for the compatible API.

Model ID

The ID of the model to use.

i Configure the research engine used when 'Enable Deep Research' is selected in features. Choose your primary tool and set parameters.

Primary Research Tool

Configures the research engine used when "Enable Research" is selected in features.

Tavily Search API Key

Used by the Tavily Search tool if enabled. Stored locally.

Research Model (OpenRouter Tool)

🔍

Search Context Size:

Low Medium (Default) High

Controls the amount of web context provided to the model (affects cost).

Max Reflection Loops

Controls iterative refinement. 0 skips reflection.

i Select the default vision-capable model used for Optical Character Recognition (OCR) and analyzing documents across all features.

Default OCR Model (for Images)

🔍

Model used for OCR and image analysis across all features.

Default Document Model (for PDFs)

🔍

Model used for native PDF analysis across all features.

i Set preferences for how new tables are formatted and where they are placed in your worksheet.

Table Format Plain Cells Formatted Excel Table (Default)

Choose how newly generated tables are added.

New Table Placement: Next Available Row Selected Cell Next Available Column (Default)

Choose where newly generated tables start.

i Control automatic styling options, like auto-fitting columns and applying background colors to newly generated cells.

Autofit Columns & Rows After Generation

i Automatically adjust column widths and row heights to fit content after new data is generated.

Applies globally whenever new content is generated.

Background Color for New Cells

Applies to plain cells, lists, and other new content.

i Manage general user experience settings, such as enabling previews before applying changes and controlling the level of progress information displayed.

Show Preview Before Applying Changes

i Display a preview of changes before applying them to the worksheet for review and confirmation.

Max Preview Rows

Max Preview Columns

Focus Preview on New/Changed Data Only

i When enabled, previews highlight only new or modified data. When disabled, shows more surrounding context.

Show Detailed Progress Information

i When enabled, displays step-by-step progress details. When disabled, shows only main status updates.

Keep progress bar after process ends

i If checked (default), the progress bar will remain visible after a process completes, allowing time to review logs. If unchecked, it will hide automatically after a short delay.

i View and manage debug logs to help diagnose issues with the add-in. You can set the log level and clear entries.

Log Level

0/Max entries (0KB) Debug logs help diagnose issues.

Log Output

Welcome to AxonSheet

Create a New Table

Add Rows to Table

Add Columns to Table

Fill Empty Cells

Edit Selected Cells

Audit & Comment Cells

Process Data Rows

Create a New Table

1 Define Your Input

Converted Document Snippet:

Deep Research Results:

2 Plan Your Output Table(s)

Add Rows to Table

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Deep Research Results:

Add Columns to Table

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Deep Research Results:

3 Plan Your Output New Columns

Fill Empty Cells

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Deep Research Output:

Smart Merge & Consolidate

1 Select Your Ranges

2 Choose Your Consolidation Goal

3 Refine with Instructions & Documents

Converted Document Snippet:

Deep Research Results:

4 Plan Your Output (Optional)

Edit Selected Cells

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Deep Research Output:

Audit & Comment Cells

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Deep Research Output:

Rows Processor

1 Select Your Range

2 Define Your Input

3 Plan Output

4 Review Input to AI

Multi Document Processor

1 Select Your Range

2 Define Your Input

Converted Document Snippet:

Import & Convert

Import Data from File

Convert JSON to Table

Generate JSON Schema

Extract Tables from Markdown

Convert Markdown to Excel

Mermaid to Diagram

Cytoscape JSON to Diagram

Add-in Settings

AI Provider & Models

OpenRouter Configuration

Ollama Local AI Configuration

OpenAI Configuration

Custom AI Service Configuration

Research Engine

Document Processing & OCR

Table Output

Styling

User Preferences

Debug Log