Getting Started


## What is Ask Steve? 
Ask Steve is a browser extension that gives your employees their own **personal copilot** which provides them with access to enterprise AI, data and software capabilities **in the flow of work**, so they can get their work done better and faster. 

## Installing Ask Steve
You can install Ask Steve in your Chrome or Edge Browser from the Webstore [here](https://quickstart.asksteve.to/browsers/).


## Working with Steve
* Once you install the Ask Steve extension you'll be dropped onto the [onboarding page](https://quickstart.asksteve.to) that will teach you how to work with Steve. 
* If you want additional information, check out the docs on the left side of this page. 
* You can pin Ask Steve to your Chrome toolbar for fast access!
<img src='/assets/images/posts/pin-extension.gif' alt="Pin Ask Steve"/>


Take your team's productivity to the next level

Welcome to Ask Steve

Learn how Steve can help you get tasks done better & faster.

Working with Steve

Skills enable you to quickly ask Steve to do complex things

Skills

Select any text or page and get help from Steve

Context Menu

Talk about life, the universe, and the current page

Chat

## What is the Writing Helper? 
The Writing Helper adds Steve to most multi-line text fields. You can ask him to write content for you or edit the content that's already in the field. It brings the power of AI right to where you work!

## The Quick Basics
* Watch the video below for a quick introduction. 
* Type whatever you want Steve to do, or pick a Skill from the list. 
* If there's no text it will help you write something. If there is, it will help you rewrite or analyze it. 
* `Enter` gives you the results in a floating panel so you can review and edit if you like. 
* `Command-Enter` (or `Control-Enter` on Windows/ChromeOS) attempts to overwrites the existing text if you're feeling lucky. This doesn't work on all sites. 
* `Command-Z` (or `Control-Z` on Windows/ChromeOS) while the Writing Helper is open tries to UNDO back to the last text. 
* Temporarily get rid of it with a right-click, permanently turn it off in the [Settings](/docs/settings). 
<br />
<div style={{ position: 'relative', paddingBottom: '56.25%', height: 0, overflow: 'hidden', maxWidth: '100%', height: 'auto' }}>
  <iframe 
    src="https://www.youtube.com/embed/ntm59m2pzgg?si=CPwEJo-JbVkDEH-6" 
    title="YouTube video player" 
    frameBorder="0" 
    allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" 
    referrerPolicy="strict-origin-when-cross-origin" 
    allowFullScreen 
    style={{ position: 'absolute', top: 0, left: 0, width: '100%', height: '100%' }}>
  </iframe>
</div>

## Bringing Up the Writing Helper
You can bring up the Writing Helper over any editable text field in the following ways: 
* All platforms: Click on the little Writing Helper icon in the upper right of any multi-line text field.
* **Mac:** Press `Command-Shift-K`
* **Windows & ChromeOS:** Press `Control-Shift-K`

If you don't like the default keyboard shortcut, you can change it! Press the puzzle piece icon in the Chrome toolbar, select `Manage Extensions` from the menu and then `Keyboard shortcuts` on the Extensions page in the left navigation menu. From there you can change the shortcuts for Ask Steve.

## Using the Writing Helper
* You can select a Skill from the dropdown menu. The Skills that will appear here are the ones that have 0 or 1 Page Variables. (See [Skills](/docs/skills))
* You can start typing in the name of a Skill and then pick it from the list. 
* Or you can type in any [one-time request](/docs/one-time-requests) if you don't already have a Skill for what you want to do. 
* Pressing `ENTER` in the Writing Helper tells Steve to do what you asked. By default Steve will put his response in a floating panel on the page. 
* If you want Steve to attempt to overwrite the existing content with his response, press `Control-ENTER` (ChromeOS, Windows) or `Command-ENTER` (Mac) instead. This doesn't work on all sites.

\
Once Steve completes your request, the Writing Helper stays open so that you can type in another request, so that you can iterate until the content is to your liking. To make it go away: 
* Click anywhere outside the box.
* Press the `Esc` key
* Press `Backspace` until it goes away

**NOTE:** If there is already text in the text field, Steve will assume that your request is about that text and will work with that text.

### Undo
If you press `Command-Z` (Mac) or `Control-Z` (Windows, ChromeOS) **while in the Writing Helper**, Steve will attempt to restore the previous text. The formatting will likely be messed up. See below on how to fix it!

### Fix Formatting
Sometimes Steve will mess up the formatting of text. You can always just ask him to `fix the formatting` and he will do so. 

## The Writing Helper icon
* This icon appears in the top right of any textarea or contenteditable text field
* If you want to make it disappear temporarily, right-click on it
* To remove it permanently from all text fields, turn it off in the [Settings](/docs/settings)

## Pro Tips
### Using a Page Variable
If the text field is empty, you can add a page variable to your request, and Steve will use that to help create the text that he puts in the text field. 
As an example, let's say you're on the page for a support case, and you want Steve to fill in an empty text field with a summary of the case. 
1. Open the Writing Helper in the empty text field
2. Write `summarize this: [[p]]` in the Writing Helper and then press `ENTER`
3. Steve will read the page, summarize it, and put the results in the text field. 
\
\
**NOTE:** This only works on empty text fields. If the text field already has text in it, Steve will read that instead. 



Writing Helper

Add permanent buttons to any page to make repetitive tasks FAST

Skill Buttons


<div style={{backgroundColor: 'violet', padding: '1rem'}}>
**The Standard Plan is limited to 1 Notebook.<br />
Premium and Enterprise Plans have unlimited Notebooks.**
</div>

\
To access the Notebooks page, click on the `Notebooks` link in the left navigation of the Extension Settings.  Or [click here](https://quickstart.asksteve.to/browsers/notebooks.html). 

## What are Notebooks?
A Notebook is a place where you can store notes as you're browsing, and then when you're ready you can visit the Notebook and ask Steve to help you analyze the notes, summarize the notes, create content based on the notes, etc. Notebooks are great for doing research, comparison shopping, and more!

## Creating a Notebook
* Go to the [Notebooks](https://quickstart.asksteve.to/browsers/notebooks.html) page in the extension Settings. 
* Press the `ADD NEW NOTEBOOK` button.
* Enter a name for your Notebook.
* Press the `SAVE` button. 

## Adding to a Notebook
* Right-mouse-click on any page or text-selection and then select `Ask Steve to...` then `Add to Notebook` and then select the Notebook you want to add to. 
* The Notebooks in this menu will be placed in alphabetical order. So you can change the order by changing the names. 
* You can use [Keyboard Shortcuts](/docs/keyboard-shortcuts) to quickly add the current page or text-selection to the first notebook in the list

## Going to a Notebook
* From the [Notebooks](https://quickstart.asksteve.to/browsers/notebooks.html) page in the extension Settings, select the Notebook from the list on the left side, then press the purple `GO TO NOTEBOOK` button. 
* From the [Context-Menu](/docs/context-menu) on any page, select `Ask Steve to...` then `Go to Notebook` and then select the Notebook you want to go to. 
* You will also get a link to the Notebook any time you add notes to it. 

## Using a Notebook
* The Notebook page is a web page like any other web page that you can use Ask Steve on. 
* So you can [Chat](/docs/chat) with the page or selection, run [Skills](/docs/skills) on the page or selection, use the [Context-Menu](/docs/context-menu), etc. 

## Editing a Notebook
* Go to the Notebook page as described above. 
* Make any edits you want, add or remove information, etc. and your Notebook will be Autosaved after 1 second. 

## Deleting a Notebook
* Go to the [Notebooks](https://quickstart.asksteve.to/browsers/notebooks.html) page in the extension Settings. 
* Select the Notebook you want to delete. 
* Press the trashcan icon and confirm deletion. 

## Where are Notebooks Saved?
They are saved on your computer. 


Quickly collect information from multiple sources so Steve can analyze it for you

Notebooks


## Why One-Time Requests?
[Skills](/docs/skills) are great for when you know ahead of time what you're going to need help with, but that's not the case a lot of the time. One-time requests enable you to ask for anything, at any time. And if you find that you're making the same one-time request over and over, you can always make a Skill!

Here are all the ways to make a one-time request with Ask Steve:
* Right-click to open the [Context-Menu](/docs/context-menu) on a page, on some selected text, or in an editable text field, and then select `Ask Steve to...` and then `🚀 Do Something`
* Press the Ask Steve toolbar icon to open the sidepanel to access [Chat](/docs/chat), and then type in your question.
* Open the [Writing Helper](/docs/writing-helper), type in your request and then press `ENTER`



All the ways that you can make a one-time request to Steve

One-Time Requests


You can change any of these shortcuts by going to chrome://extensions in your browser and then clicking on "Keyboard shortcuts" in the left navigation menu. 

## Mac
* `Command-Shift-K`: Brings up the [Writing Helper](/docs/writing-helper) when pressed in any text field where you can type.
* `Command-Shift-X`: Adds the current page or text selection to the 1st [Notebook](/docs/notebooks) in your Notebook list.
* `Command-Shift-1`: Runs the first pinned [Skill](/docs/skills). [video](https://youtu.be/bJHYI-AhMz4) 
* `Command-Shift-2`: Runs the second pinned [Skill](/docs/skills). [video](https://youtu.be/bJHYI-AhMz4)

## Windows, ChromeOS
* `Control-Shift-K`: Brings up the [Writing Helper](/docs/writing-helper) when pressed in any text field where you can type.
* `Control-Shift-X`: Adds the current page or text selection to the 1st [Notebook](/docs/notebooks) in your Notebook list.
* `Control-Shift-1`: Runs the first pinned [Skill](/docs/skills). [video](https://youtu.be/bJHYI-AhMz4)
* `Control-Shift-2`: Runs the second pinned [Skill](/docs/skills). [video](https://youtu.be/bJHYI-AhMz4)


Keyboard Shortcuts


<div style={{backgroundColor: 'violet', padding: '1rem'}}>
**This is only available on the Premium Plan.**
</div>
<br />
<iframe width="560" height="315" src="https://www.youtube.com/embed/4879AXfUaLo?si=EVu_4J5okgUIZ_YQ" title="YouTube video player" frameborder="0" allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share; fullscreen" referrerpolicy="strict-origin-when-cross-origin" allowfullscreen></iframe>

## Why Call an API or Webhook? 
* This gives you the ability to call your own APIs that can do pre- and post-processing. 
* As an example, you could configure your own internal API that Steve calls with an account number that he pulls from a page. Your API can use that account number to pull relevant data about that account from internal systems, directly make an LLM call with the new augmented data, and then pass the result back to Steve. 
* Or you could send the input to an LLM for processing, and then write it out to a Google Sheet or to Salesforce. 
* Or you could proxy all requests so that you can log them, or filter any sensitive information out. 
* You can do this with your own code running on a server somewhere, or with any visual workflow builder like [n8n](https://n8n.io/), [Pipedream](https://pipedream.com/), [Activepieces](https://www.activepieces.com/), [Make](https://www.make.com/), [Zapier](https://zapier.com/), [Gumloop](https://www.gumloop.com/) etc. that supports Webhooks. 
* This enables you to do a lot more than just call an LLM! 

<div style={{backgroundColor: 'lightgoldenrodyellow', padding: '1rem', marginTop: '20px'}}>
**Check out [this Blog Post](/blog/tutorials/make-lead-to-sheets) for a detailed walkthrough showing how to connect Ask Steve to a Make.com API**
</div>

## How to Call an API or Webhook
Ask Steve can interface with any API or Webhook on the back-end. 
* To point a Skill to an API/Webhook instead of an LLM, follow the instructions [here](/docs/skills#advanced-options) when setting up your Skill.
* API/Webhook Skills behave as follows: 
1. All requests are made via POST
2. In the Skill configuration, the default header is to set the Content-Type to "application/json", but you can add anything else to the header that you need, including an API key or Authentication key. 
3. The body of the POST request contains a property for every [Page Variable](/docs/skills#page-variables-in-skills) and [User Variable](/docs/skills#user-variables-in-skills). All other text in the "Instructions for Steve" area are ignored. You can also include additional static variables in the body in the Skill configuration. 
5. The default behavior is that your API/Webhook should return an object with a "message" property (string) if it was successful, and an "error" property (string) if it was not. This is what will be shown to the user. You can change this if desired in the Skill configuration. 
6. You can include simple HTML formatting in the return values. URLs will be hyperlinked automatically, and text can be hyperlinked by using [Markdown formatting](https://anvilproject.org/guides/content/creating-links). 

### System Variables
The following System Variables can optionally be sent to your API or Webhook if you include them in your instructions in single brackets, e.g. [_url], [_html], [_timestamp].
1. **_html**: has a stripped down version of the HTML of the current page. Most non-text nodes are removed. This is helpful if you need to do some analysis of the page structure. 
1. **_language**: has the language of the browser making the API call, as returned by navigator.language
1. **_os**: has the operating system of the computer making the API call, as returned by [chrome.runtime.getPlatformInfo](https://developer.chrome.com/docs/extensions/reference/api/runtime#method-getPlatformInfo)
1. **_preferences**: has the user's [Preferences](/docs/preferences). 
1. **_prompt**: has the full prompt - so all the text in the Skill with all the variables replaced, as if you were sending a single prompt to an LLM. All page and user-input variables will be replaced, and all System Variables will be removed. 
1. **_skillName**: has the name of the Skill being called.
1. **_timestamp**: has the timestamp of the API call. This is a numeric value (integer) representing the number of milliseconds since the Unix epoch, as returned by the Javascript Date.now() function. 
1. **_title**: has the title of the current page. 
1. **_url**: has the URL of the current page. 
1. **_userAgent**: has the user agent of the browser making the API call, as returned by navigator.userAgent



> **Example Instructions:** *This Skill captures [[leads]] and sends them to a Webhook. [_url][_title]*
>
> The POST body will have the specified property "leads" with the selected text or current page. The system variables "_url" and "_title" will also be added. All other text is ignored and not sent to the API/Webhook. 

> **Example Instructions:** *This Skill captures a [[headline]], [[author]] and [[article]] then asks the user for \{\{what do you want to do with this\}\} and then sends them to a Webhook*
>
> The POST body will have the three specified page variables, "headline", "author", "article" and the user-input variable "what do you want to do with this". All other text is ignored and not sent to the API/Webhook. 


Calling APIs and Webhooks


Ask Steve formats the results from the LLM or API in the following ways:

## Hyperlinks 
* Any URLs that aren't the `src` or `href` of an HTML tag are automatically hyperlinked.

## Markdown
* Any text in surrounded by two asterisks like `**this**` is bolded
* Markdown links of the format `[Link](https://link.to.website)` are turned into Hyperlinks
* Code blocks like ```html are wrapped in `<pre><code>` tags

## HTML
* Most HTML should be supported. 
* All `<img>` tags are linked to the source image. 



Formatting Results


Ask Steve can work on your local (downloaded) files via our [File Chat](https://quickstart.asksteve.to/viewer) page. This page extracts the text out of your files so that Ask Steve can use it. 

## Accessing File Chat
* You can open File Chat from the bottom of the Ask Steve [Context Menu](/docs/context-menu)
* Also from the sidebar in the extension [Settings](https://quickstart.asksteve.to/browsers/options.html)
* Or by clicking [here](https://quickstart.asksteve.to/viewer/)

## Using File Chat
* Click where it says "Click here to select a file" to pick a file. 
* Alternately you can drag and drop a file onto the purple dropzone. 
* You can also paste an image file into this page and it will be processed.
* You can also paste any text into this page. 

## Supported File Types:
* pdf, docx, pptx, jpg, png, gif
* For PDFs, the page tries to extract any text in the file. If it can't find any, it turns each page into an image and then OCRs it. 

## Processing
* All the processing is done in your browser. Your files never leave your computer. 



File Chat

Settings


## Accessing the Ask Steve Login page
You can access the Login page in the following ways: 
* Right-click on the Ask Steve toolbar icon and select `Options`
* Click on the Ask Steve toolbar icon or lightning bolt tab to open the sidepanel, then press the gear icon in the upper right
* Following [this link](https://quickstart.asksteve.to/browsers/options.html)

## Sign Up / Login
* Here's where you can see your current plan and can create an account or login. You'll need to login in order to upgrade to Standard or Premium. 
* You can create an account with an email and password, in which case a confirmation email will be sent to you. 
* You can also login with your Google account, in which case no email confirmation is needed. 

## Setting Up a Free Model
* Once you login you can add your API Key for a free model. This will switch you from the QuickStart plan to the Standard plan.
* You can get your **FREE** Google Gemini API Key by following the instructions [here](https://quickstart.asksteve.to/get-gemini-key.html). **Free Gemini is not available in the EU, UK or Switzerland** ([details](/docs/where-it-works#countries)).
* You can get your **FREE** Mistral La Plateforme API Key by following the instructions [here](https://quickstart.asksteve.to/get-mistral-key.html). **This is available worldwide.**
* Sign up for the Premium Plan if you'd like to connect to other public models (OpenAI, Cohere, Anthropic, etc.), an on-premise model, a local model running on your own computer, or to a non-model API.



Login to Ask Steve and set up your first model

Login


To access the Settings page, click on the `Settings` link in the left navigation of the Extension Settings.  Or [click here](https://quickstart.asksteve.to/browsers/settings.html). 

## Change the Ask Steve Sidepanel Theme
* **Light:** Use the light theme
* **Dark** Use the dark theme
* **System:** Use whatever theme your OS is set to use

Whenever you close and reopen the sidepanel, it will check for the latest theme setting and use it. 

## Show the Ask Steve Tab
<img src='/assets/images/docs/tab.png' alt="Tab"/>
* By default Ask Steve shows a tab on every page that gives you quick access to the Ask Steve sidepanel. It's purple with a lightning bolt. 
* Pressing the button opens the Ask Steve sidepanel.
* You can grab the drag handle on the tab up and down the right edge of the page and Steve will remember where you placed it.
* If you'd like to turn off the tab, uncheck the `Show Tab` checkbox and press `SAVE`
* If you'd like to temporarily hide the tab because you need to take a screenshot or it's just in your way, right-click on it. It will disappear from the current page until the page is reloaded. 

## Show the Ask Steve Writing Helper
<img src='/assets/images/docs/textbug.png' alt="Writing Helper"/>
* By default Ask Steve shows a little icon in every textarea that gives you quick access to writing and rewriting help with the [Writing Helper](/docs/writing-helper).
* Pressing the icon opens the Writing Helper.
* If you'd like to turn off the icon, uncheck the `Show Writing Helper` checkbox and press `SAVE`
* If you'd like to temporarily hide the icon because you need to take a screenshot or it's just in your way, right-click on it. It will disappear from the current textarea. 




To access the Skill Library page, click on the `Skill Library` link in the left navigation of the Extension Settings. Or [click here](https://quickstart.asksteve.to/browsers/library.html).

## Installing a Skill
* Click the tile for any Skill that you're interested in. You can review the author's notes and the instructions for Steve.
* If you want to install the Skill, click on the `INSTALL` button.
* The Skill will be added to your Skills list. If you want to edit it, go to [Skills](/docs/skills) tab in the Ask Steve sidepanel. 

\
**NOTE:** Existing Skills are overwritten, so if you have made any customizations, they will be lost.


Find and install new Skills, then customize them and make them yours!

The Skill Library

Import, export and delete your Skills and Skill Buttons

Import, Export & Delete


To access the Preferences page, click on the `Preferences` link in the left navigation of the Extension Settings.  Or [click here](https://quickstart.asksteve.to/browsers/preferences.html). 

## What are Preferences?
* You can tell Steve facts or instructions that you want him to always keep in mind. 
* An example of a fact might be *My name is Inigo Montoya and the six-fingered man killed my father*
* An example of an instruction might be *Always respond in Spanish*
* An example of something more involved might be your whole resume.
* Every request you make to Steve includes your preferences. This way you can tell Steve something once and he'll "know" it every time he helps you out.  

> For example: If I save a memory with my name, job title and company, then in any Skill I can ask Steve to use them and he will know what they are. 

## Editing your Preferences
* On the [Preferences Page](https://quickstart.asksteve.to/browsers/preferences.html) update your preferences and then save them. 

## Pro-Tips
* Try to keep your preferences on the shorter side, as they take up some of the limited room in your LLM request
* Preferences don't always work like you'd expect - some models are better than others at using them. You will need to experiment!




Have Steve remember facts and instructions about you

Preferences

<div style={{backgroundColor: 'violet', padding: '1rem'}}>
* **This is only available on the Premium Plan.**
* **You need to have your own account (paid or otherwise) with the LLM service/model provider.**
</div>


To access the Models page, click on the `Models` link in the left navigation of the Extension Settings. Or [click here](https://quickstart.asksteve.to/browsers/models.html).

## What Does the Model Do? 
* The model is the LLM service that Steve uses to help you accomplish all your tasks. 
* During your QuickStart trial Ask Steve uses Google's Gemini 1.5 Flash model.
* After that, you can use the same model with your own API Key, or Mistral's free La Plateforme model.
* All models have different characteristics and are better or worse at various tasks.
* Generally the better the model, the slower and more expensive it is. 
* **When you use the free Gemini or Mistral models, those companies reserves the right to use your data for training purposes and to make their products better.** 
* If your data is sensitive or you are concerned about privacy, you should sign up for a Premium Ask Steve account and then connect Ask Steve to an account at a provider that doesn't use your data. 
* Switching out Steve's model is like switching out his brain. 

## QuickStart Model
* The QuickStart plan uses Google's Gemini 1.5 Flash model.
* When you are on the QuickStart plan, your requests are sent to the Ask Steve server, which sends them along to Google, and then sends the results back. Your requests go through our server, but we never save anything. 

## Add a New Model
* To add a new model, press the `ADD NEW MODEL` button. A dialog will open with a number of model templates that you can select from, and a place for you to enter your API Key for that model if you have it already. Press `OK` and the interface will be filled in with the template details. 
* Edit any of the details that you need to
* To test the model, scroll down to the `Test Your Connection` area and press the `TEST` button. A sample query will be sent and the response shown below. 
* Once you are satisfied that everything is working correctly, press the `SAVE NEW MODEL` button. Steve will switch to using this new model, as indicated in the dropdown list at the top of the page. 

All requests are made via POST. The data fields for a model are as follows: 
* **Name:** So that you know what model this is. This is shown to users in error messages. 
* **URL:** The URL endpoint for this model
* **Context Window:** The size of the context window for this model. This can usually be found on their documentation pages. If in doubt, put in 32000
* **Header:** The JSON header that is sent in the POST request. This should be formatted according to the service provider's documentation. This typically includes the API Key or other authentication for the service. 
* **Body:** The JSON body that is sent with the POST request. This should be formatted according to the service provider's documentation. 
* **Response Path:** In bracket notation, the path to the field in the response object that contains the actual response. This will be in the service provider's documentation. Note that one-shot responses and streaming responses usually have different paths, even from the same service provider. 
* **Error Path:** In bracket notation, the path to the field in the response object that contains any error messages. This will be in the service provider's documentation.
* **This is a Streaming API:** If you are calling a streaming API, this needs to be checked, as streaming responses are parsed differently than non-streaming responses. 

## Edit a Model
* Select the model that you want to edit from the `Steve is using:` dropdown. 
* Make your edits
* Press the `UPDATE` button
\
\
**NOTE:** This also changes the current model. Change it back when you're done editing via the `Steve is using:` dropdown. 

## Change the Current Model
* Select the model that you want to change to from the `Steve is using:` dropdown. 
* This is automatically saved, so now Steve will use this model.


## Delete a Model
* Select the model that you want to delete from the `Steve is using:` dropdown. 
* Press the `DELETE` button and confirm deletion
* Select whatever model you want Steve to use now

## Pro-Tips
### Call Any API
* If you want to call an API or Webhook from a Skill, see the details [here](/docs/webhooks)

### Change the Temperature
* The `temperature` setting of a model usually ranges from 0.0 to 1.0 or 2.0, where 0 tells the model to provide more factual answers, and the upper limit tells the model to provide more creative responses. The value can be set to a value somewhere in that range (e.g. 0.3)
* The default model and all our templates use a temperature of 0, as most use cases will benefit from more factual responses. 
* When adding a new model, you can set its temperature to something else, depending on how you're going to use it. 
* You can even add the same model multiple times, each with a different temperature, and then assign Skills that need more creativity to higher temperature models. 

### Connect to a Model Running On Your Own Machine
* See details on how to set this up [here](/docs/local-models)

Connect to other models where you have accounts

Connecting to Other Models

<div style={{backgroundColor: 'violet', padding: '1rem'}}>
**This is only available on the Premium Plan.**
</div>

## What is a Local Model?
When you run a local model, the LLM runs on your own computer. Then Steve can use that model instead of using one in the cloud. The primary benefit of this is that **your requests never leave your machine, so all your data is kept private**. The secondary benefit is that it's FREE. We have tested Ask Steve with [LM Studio](https://lmstudio.ai/) and [Ollama](https://ollama.com/), but you should be able to use any system that can expose a model via a local web server.

### What are the downsides? 
* You have to download big model files (e.g. 3+GB)
* You need to have a fast machine
* The models your local machine can run generally aren't as good as the ones you can access in the cloud
* It can be slower, even though it's running locally

## How Do I Use LM Studio With Ask Steve?
1. Download and setup [LM Studio](https://lmstudio.ai/)
2. Open LM Studio. From the home page download a model. Try `Llama 3 - 8B Instruct`.
3. Once it's downloaded, go to the [Local Server](https://lmstudio.ai/docs/local-server) page in LM Studio and select the model you just downloaded from the dropdown at the top. If it doesn't start automatically, press the green Start Server button. This will start up a local server on port 1234 with the selected model.
4. Finally, go to the [Models](https://quickstart.asksteve.to/browsers/models.html) page in Ask Steve Settings, press `ADD NEW MODEL` and select `Local: LM Studio` from the menu. You won't need an API Key since it's running on your own machine. Press `TEST` to ensure it works, then `SAVE NEW MODEL` to save it. 
5. Congratulations! Steve is now using a model running completely on your own computer!


## How Do I Use Ollama With Ask Steve?
1. Download and setup [Ollama](https://ollama.com/)
2. Download a model. For example, this terminal command will pull Meta's Llama 3: `ollama pull llama3`
3. You will need to enable the Ask Steve extension to connect to Ollama. To do this you need to configure the Ollama server with the environment variable OLLAMA_ORIGINS, and it needs to be set to "chrome-extension://gldebcpkoojijledacjeboaehblhfbjg”. For Microsoft Edge it should be "chrome-extension://hiefenciocpoafbgoocochpolfjfmjfg"
4. Instructions for how to do so on various platforms are [here](https://github.com/ollama/ollama/blob/main/docs/faq.md#how-do-i-configure-ollama-server). So on Chrome on a Mac you'd issue this terminal command: `launchctl setenv OLLAMA_ORIGINS "chrome-extension://gldebcpkoojijledacjeboaehblhfbjg"` to give Ask Steve access to Ollama. 
5. After setting OLLAMA_ORIGINS, you will need to restart the Ollama server. On a Mac you can do this by quitting and restarting Ollama from the Mac Taskbar. 
6. Finally, go to the [Models](https://quickstart.asksteve.to/browsers/models.html) page in Ask Steve Settings, press `ADD NEW MODEL` and select `Local: Ollama - Meta Llama 3` from the menu. You won't need an API Key since it's running on your own machine. Press `TEST` to ensure it works, then `SAVE NEW MODEL` to save it. Note that this assumes you downloaded the Llama 3 model. Change the "model" parameter in the body if you downloaded a different model. 
7. Congratulations! Steve is now using a model running completely on your own computer!




Connect to a model running on your local machine via Ollama or LM Studio

Connecting to a Local Model

Other



## Websites are very... variable
Ask Steve will work on most websites, but given how powerful the web platform is and all the different ways that people can build sites and apps, it doesn't work everywhere. 
* It will most likely not work in web applications with complex layouts
* Sites that make excessive use of iframes and Shadow DOMs may cause problems

### Known Sites Steve Has... Issues
* Figma
* XHTML sites, like the [SEC's document viewer](https://www.sec.gov/ix?doc=/Archives/edgar/data/1652044/000165204424000022/goog-20231231.htm). On sites like this, Steve does what you ask him, but then can't write his response to the page. You can always see his response by opening Chrome DevTools (Chrome > View Menu > Developer Tools) and then going to the Console tab and looking for the text where Steve is waving 👋. 
* Salesforce: When sending an email, Salesforce doesn't "see" any of Steve's changes until you manually edit something in the email. So always be sure to edit Steve's work! 

### Chrome Security Policies
* If you're using Ask Steve at work, your company may have set policies in Chrome that prohibit either the installation of certain extensions or the use of extensions on specific URLs. 
* In this case, you would need to ask a company administrator to add Ask Steve to the allowlist. 
* You can see what policies your browser is enforcing by opening `chrome://policy` and then looking for `Extension Settings`


## Browsers
Since Ask Steve is built for Chrome, it will at least partially work in most other Chromium-based browsers.
* **Chrome:** Works great. Install from the [Chrome Web Store](https://chromewebstore.google.com/detail/ask-steve-add-ai-superpow/gldebcpkoojijledacjeboaehblhfbjg?hl=en&authuser=0)
* **Edge:** Works great. Install from the [Edge Add-Ons Store](https://microsoftedge.microsoft.com/addons/detail/ask-steve-add-ai-superp/hiefenciocpoafbgoocochpolfjfmjfg)
* **Vivaldi:** Mostly works. The FAB to open/close the sidepanel does not. Install from the [Chrome Web Store](https://chromewebstore.google.com/detail/ask-steve-add-ai-superpow/gldebcpkoojijledacjeboaehblhfbjg?hl=en&authuser=0)
* **Brave:** Untested
* **Opera:** Untested
* **Arc:** Does not support the sidepanel so you cannot [Chat](chat) with a page or create or edit [Skills](skills) or [Skill Buttons](skill-buttons). You can still use the [Context Menu](context-menu) and the [Writing Helper](quick-action-box). You can try to use a supported browser to create/edit your Skills and Skill Buttons, [export](export#export) them and then [import](export#import) them into Arc. Install from the [Chrome Web Store](https://chromewebstore.google.com/detail/ask-steve-add-ai-superpow/gldebcpkoojijledacjeboaehblhfbjg?hl=en&authuser=0)
* **Safari:** Not supported
* **Firefox:** Not supported

Ask Steve works on most websites, but not all of them

Where Does Ask Steve Work?


## Your Privacy
We take your privacy very seriously.
* We don't track what websites you visit
* We don't track what you're asking Steve for
* We don't track the Skills that you create
* We don't store your API Keys on our servers
* You can read our full privacy policy [here](/privacy).

## Our Philosophy
* Respect the user's privacy (see above)
* Try to keep Ask Steve servers out of the loop as much as possible
* Store as much data on the user's machine as possible
* Connect directly to the LLM provider whenever possible
* Be as transparent as possible about your data, how we use it, and where we send it (see below)
* Ask Steve was created by [Rajat Paharia](https://www.linkedin.com/in/rajatrocks/) who formerly worked at Google and IDEO. He created it because he wanted to put control of AI into your hands instead of the hands of big companies. See more [here](/about).


## Extension Permissions
* When you install Ask Steve it will ask for permission to *"Read and change all your data on all websites"*. It needs this permission in order to be helpful on any web page. If you are uncomfortable with this then you should press `Cancel` and not install the extension. 
* If you want to limit what sites Ask Steve can work on, you can go into `chrome://extensions`, press the `Details` button on the Ask Steve tile, and then under `Site access` control what sites you want to give Steve access to.
* At any time you can disable Ask Steve via the toggle on the Manage Extensions page and enable it manually on sites you want to use it on.
* In an enterprise deployment, your administrator can configure certain sites to be allowed/disallowed via the `runtime_allowed_hosts` and `runtime_blocked_hosts` keys in [ExtensionSettings](https://support.google.com/chrome/a/answer/9867568?hl=en). You can use this to prevent Ask Steve from having access to sensitive internal pages. 
* The Ask Steve extension id on Chrome is: **gldebcpkoojijledacjeboaehblhfbjg**
* The Ask Steve extension id on Edge is: **hiefenciocpoafbgoocochpolfjfmjfg**

## Saved Data
### Your API Keys
* Ask Steve asks you to enter your LLM provider API keys so that it can connect to those providers.
> These keys are stored in the extension's [Local Storage](https://developer.chrome.com/docs/extensions/reference/api/storage) on your computer and are not sent to Ask Steve's servers.
* The API keys are only sent to the service provider in order to authenticate your requests. 
* The API Keys are not encrypted and are sent with every request to the service provider, so someone else with access to your computer could discover them. 

### Other Data
All other configuration data is stored in Local Storage on your computer. It is never stored anywhere else. This includes: 
* Skills
* Skill buttons
* Model configuration
* Chat history

> If you'd like to see what is being stored in Local Storage, do the following: 
> * Go into `chrome://extensions`
> * Enable `Developer mode` via the toggle in the upper right
> * Press the `service worker` link in the Ask Steve tile. This will open Chrome DevTools
> * On the Console tab, paste `await chrome.storage.local.get()` and then press `ENTER`
> * The contents of Local Storage will be printed out and you can inspect them

## Network Requests
### Your LLM Requests
* We send your request directly to the LLM service provider. 
* When you are on the QuickStart plan, your LLM requests are sent to to **quickstart.asksteve.to** which proxies them to Google's Gemini or another provider. The QuickStart proxy doesn't look at or store your request, it just passes it through. 

### Google Analytics
We log views of the extension pages (e.g. Settings, Models, Sidepanel, etc.), as well as the following information for every LLM request:
* Whether the call was successful or returned an error.
* The hostname for the service (e.g. `generativelanguage.googleapis.com` or `api.openai.com`)
* A transient session identifier
* This can be disabled for enterprise deployments
* Here is what that looks like for a call to the Google Gemini LLM:
```
{
  "name": "llm_success",
  "params": {
    "hostname": "generativelanguage.googleapis.com",
    "session_id": "1721321767437",
    "engagement_time_msec": 100
  }
}
```

### Login and Authentication
* Ask Steve uses Firebase (from Google) for login and authentication. Requests are made to Google's servers whenever you login or create a new account. 
* We don't ask for or store any PII beyond what is provided at login. 
* This can be disabled for enterprise deployments.

### The Skill Library
* Ask Steve uses Firebase for its backend database. Requests are made to Firebase to retrieve the list of Library Skills and to (in the future) upload a Skill to the Library.
* This can be disabled or pointed to an internal database for enterprise deployments. 

### Configuration Files
Ask Steve loads the following configuration files from **quickstart.asksteve.to**
* **models.json** - contains all the model templates
* **skills.json** - contains the out-of-the-box Skills
* **buttons.json** - contains the out-of-the-box Skill Buttons
* This can be disabled for enterprise deployments. 


> If you'd like to monitor all the network requests that Ask Steve is making, do the following: 
> * Go into `chrome://extensions`
> * Enable `Developer mode` via the toggle in the upper right
> * Press the `service worker` link in the Ask Steve tile. This will open Chrome DevTools
> * Switch to the Network tab. From here you can see all the network requests that Ask Steve makes

## Other
The LLM service may use your data for training or to make their products better. If you are on the QuickStart plan or are using the Google Gemini or Mistral La Plateforme free tier, they reserve the right to use your data. If this concerns you, you should sign up for Ask Steve's Premium Plan and connect to a paid service or internal corporate service that explicitly states that they will not use your data. You can also connect to a [Local Model](/docs/local-models) running on your own machine. 






What permissions does Ask Steve ask for and how does it use your data?

Permissions & Your Data


## Standard (Free!)
* When you first install Ask Steve you are put on our QuickStart trial. This gives you 100 requests that you can use to try out working with Steve and see if you find him helpful.  Responses from Steve come all at once.
* Ask Steve asks you to enter your own free Google Gemini or Mistral La Plateforme API Key to enable the Standard plan.
* [See how to get your own Google Gemini key here](https://quickstart.asksteve.to/get-gemini-key.html). 
* [See how to get your own Mistral La Plateforme key here](https://quickstart.asksteve.to/get-mistral-key.html). 
* You continue to have access to all the functionality, with the only limits being what service provider imposes. The free Google Gemini API Key has [the following limits](https://ai.google.dev/pricing). The Mistral La Plateforme free tier has [the following limits](https://console.mistral.ai/limits/).
* **Free Gemini is not available in the EU, UK or Switzerland** ([details](/docs/where-it-works#countries)). If you are in one of those regions, you should use Mistral's La Plateforme. 
* Steve's responses are streamed, so you can work even faster.  
* Both of these free offerings reserve the right to use any data that you send them for training. **If you don't want your data used for training, consider signing up for the Premium Plan.**
* Community support

## Premium
<div style={{backgroundColor: 'violet', color: 'black', padding: '1rem', marginBottom: '10px'}}>
<a href="https://asksteve.to/pricing" style={{fontWeight: 'bold', color:'black'}}>Upgrade to Premium</a>
</div>

* [Connect to any paid model](/docs/models)
* [Connect to any on-premise model](/docs/models)
* [Connect to any local model](/docs/local-models)
* [Connect to any API](/docs/webhooks)
* [Assign specific models to specific Skills](/docs/skills#advanced-options)
* Email support



## Enterprise (Coming Soon)
* Central team billing
* Central team user management
* Central team Skill Library
* Custom Skill packs (like in [Onboarding](https://quickstart.asksteve.to)) for your users
* Backup to server
* Integration services
* On-premise deployment
* Email support


Plans


> Want to revisit the onboarding pages? [Follow this link!](https://quickstart.asksteve.to)

Want to run through the onboarding again?

Onboarding Pages


We would appreciate any support that you can give us so that we can continue to make Ask Steve even better! Here are some ways that you can help us: 
* Signing up for our [Premium Plan](/pricing)
* Writing a good review on the Chrome or Edge Web Store [here](https://quickstart.asksteve.to/browsers/)
* Sending us a testimonial about how you're using Ask Steve [here](https://forms.gle/KAEAeCLsGeLAGqKk8)
* Spreading the word and telling your friends and coworkers about Ask Steve
* Sharing tips, tricks and Skills in our [User Community](https://groups.google.com/g/ask-steve)





Support Us


## v1.2.4 - December 18, 2024
* Added a [Writing Helper](/docs/writing-helper) icon to textareas and other contenteditable controls. Click on it to get writing or rewriting help in the control. Press `ENTER` to get the results in a floating panel, or `COMMAND-ENTER` (Mac) or `CONTROL-ENTER` (PC, ChromeOS) to write over the existing text.
* Temporarily hide the Writing Helper icon by right-clicking on it. 
* Permanently hide the Writing Helper icon on the [Settings](/docs/settings) page. 
* Added Shortcut buttons to the Tab that opens the Ask Steve sidepanel. Hover over the tab for a second and buttons to quickly trigger your top 2 pinned Skills will appear.
* Fixed minor bugs 


## v1.2.3 - December 9, 2024
* Added the ability to have multiple and single selection controls when requesting [User Input](/docs/skills#multiple-choice-in-user-variables). 
* Updated the User Input UI
* Can now `Command-ENTER` (Mac) or `Control-ENTER` (Windows, ChromeOS) to submit the User Input form. 
* You can now return img tags in responses. The images are embedded and a link to open the image in a new window is automatically added.
* You can now resize the floating panel and it remembers the last size.
* Updated the [Skill Library](/docs/library)
* Remove all extraneous control characters from content submitted to LLMs and APIs. 
* Fixed minor bugs


## v1.2.2 - November 26, 2024
* Fixed minor UI bugs


## v1.2.1 - November 25, 2024
* Changed the keyboard shortcuts to be more useful ([video](https://youtube.com/shorts/wOdr-Q7jImc?feature=share)):
* `CMD/CTRL-SHIFT-1` runs the first pinned Skill on the current page or selection. Pinned Skills are in alphabetical order. It used to capture the page or selection into the first [Notebook](/docs/notebooks). 
* `CMD/CTRL-SHIFT-2` runs the second pinned Skill on the current page or selection. Pinned Skills are in alphabetical order. It used to capture the page or selection into the second Notebook. 
* `CMD/CTRL-SHIFT-X` captures the current page or selection to the first Notebook. It used to open/close the Ask Steve sidepanel.
* `CMD/CTRL-SHIFT-K` has no change - it opens the [Writing Helper](/docs/writing-helper) in an editable text field
* You can change any of these shortcuts or add your own open/close shortcut by going to `chrome://extensions/shortcuts` and adding a shortcut for "Activate the extension".
* Added support for [Google Apps Script](https://youtube.com/shorts/f0UEc401acg?feature=share)
* Added support for [PDFs](https://youtube.com/shorts/OKJY9IDQrQY?feature=share) without having to download the PDF first
* Changed PDF Chat to [File Chat](https://quickstart.asksteve.to/viewer) which now supports pdf, docx, pptx, jpg, png, and gif, as well as pasting an image or text. ([video](https://youtube.com/shorts/gol7lFHSw1E?feature=share))
* Added support for [YouTube videos](https://youtube.com/shorts/CW_HoT3qKb8?feature=share)
* Better code formatting in Chat
* Skill Buttons now use selected text (if there is any) instead of the whole page
* If you start a Skill Button name with an emoji, the default ⚡ will be removed
* Added "Open File Chat" to the context menu so you can get there quickly from anywhere
* Created a [Google Apps Script Skill Pack](https://quickstart.asksteve.to/)


## v1.2.0 - November 16, 2024
* Added support for Chatting with Google Docs. See demo video [here](https://youtube.com/shorts/hHxU7_81seM?feature=share)
* Added support for Chatting with Google Slides. See demo video [here](https://youtube.com/shorts/-HUWHBDWp14?feature=share)
* You can Chat with a selection in Microsoft Word or PowerPoint online, but not a whole document. 
* Added a [PDF Chat tool](https://quickstart.asksteve.to/viewer) where you can pick a PDF or image file and have it turned into text so you can use it with Ask Steve. See demo video [here](https://youtube.com/shorts/3QhSskiZVoc?feature=share)
* Chat now works on blank tabs.
* Added the ability to add static body variables and specify the result and error paths for an API call. See updated docs [here](/docs/webhooks). Removed special-purpose code for Zapier, since now you can specify ["status"] as the result path.
* Sidepanel no longer totally resets when you switch tabs. 
* Other minor updates


## v1.1.9 - November 10, 2024
* Added support for [Zapier webhooks](https://zapier.com/blog/what-are-webhooks/). Zapier webhooks return right away with a status of "success" and then process asynchronously. So you can trigger Zapier workflows from Ask Steve, but cannot return any data. 
* Ask Steve is now live on the [Edge Add-Ons store](https://microsoftedge.microsoft.com/addons/detail/ask-steve-add-ai-superp/hiefenciocpoafbgoocochpolfjfmjfg)
* Updated all store copy and assets


## v1.1.8 - Skipped

## v1.1.7 - October 28, 2024
* Added a new [System Variable](/docs/webhooks#system-variables) for API calls/webhooks: [_imageUrls] sends an array with the URLs of all the images on the page. 
* Made the z-index of Skill Buttons higher so they don't get hidden on pages
* Changed the top menu item to "Do Something" instead of "Ask me what to do"
* Changes in plan are now detected automatically and don't require logging out and back in
* Added a link to the extension settings sidebar to manage your subscription
* Notebooks now show linebreaks from the original page
* Other minor updates


## v1.1.6 - October 14, 2024
* Added several more optional [System Variables](/docs/webhooks#system-variables) to API calls/webhooks to enable more use cases. "_url" used to be automatically included but now needs to be explicitly added.
* Added new [Notebooks](/docs/notebooks) functionality.
* Added new [Preferences](/docs/preferences) functionality. 
* Removed old Memories functionality. 
* Added more [Keyboard Shortcuts](/docs/keyboard-shortcuts)
* Updated the context-menu. Changed the "ONE-TIME REQUEST" menu item to "Ask me what to do", added separators, etc. 
* Other minor updates. 


## v1.1.5 - September 24, 2024
* Implemented the Plans described on the [Plans and Pricing](/pricing) page. **You now need a paid Premium plan to use any model aside from the free Gemini or Mistral. You also need a Premium plan to use APIs and Webhooks.**
* The UI has been changed to reflect this. 
* Updated the model templates for Anthropic and Mistral.  
* Added an "Add Skills" button to the top of the Ask Steve sidepanel to make it easy to access the Skill Library. 
* Removed the error messages that were showing when you tried to chat on Google Docs, Slides or Sheets
* Other minor updates. 
 
## v1.1.4 - Skipped

## v1.1.3 - September 16, 2024
* Right click on the FAB/TAB (the purple tab with the lightning bolt icon) to temporarily remove it from a page, in case you need to print or take a screenshot. 
* Removed scrollbars on Windows on the Import/Export page.
* Updated the confirmation dialogs for adding Skill Packs.
* Other minor updates.  


## v1.1.2 - August 27, 2024
* The Ask Steve lightning bolt tab/FAB no longer appears when you print pages. 
* Fixed issues with the format of text generated by the [Writing Helper](/docs/writing-helper) in Outlook 365. 


## v1.1.1 - August 16, 2024
* Many minor bug fixes
* Added better error messages if you try to use Ask Steve with Google Docs, Slides, Sheets or Figma, since it can't read content from any of those. 


## v1.1.0 - July 30, 2024
* Fixed bugs.  
* Changed how calls to APIs/Webhooks are made to make it easier - see the updated [documentation](/docs/webhooks). 
* As a result, removed the "Show Model Override" setting on the Settings page and the "This is Not an LLM" setting on the Models page since they're no longer needed. 
* Removed omnibox support


## v1.0.9 - July 24, 2024
* Automatically hide the FAB when a browser window is in fullscreen mode (so it doesn’t get in the way of your YouTube videos)
* Chat and the Floating panel now hyperlink all URLs
* Added a new “Show Model Override” setting on the Settings page that adds the ability to assign a specific model to a specific Skill. This is useful for when certain models are better at certain tasks, but also for when you want to call non-model APIs. See the video in the next bullet for an example.
* Added a “This is Not an LLM” setting on the Models page that indicates that you’re calling a non-LLM API. All the usual instruction text and Memories that Steve would send with your request are removed. See [this video](https://www.youtube.com/watch?v=uUXwVwdVUgE) for an example. 



## v1.0.8 - July 18, 2024
* OpenAI [made a change](https://community.openai.com/t/streaming-events-returned-bunched-up/870324) to how they send down streaming responses that broke our parsing. This has been fixed. 
* The sidepanel code has been changed so that there is one universal sidepanel for all the tabs in a window vs. a unique sidepanel per tab.
> * This change was made because unique sidepanels per tab is not supported in Edge and Vivaldi, and we weren’t really doing anything with that capability anyway. 
> * So now when you open the sidepanel, it is open on all the tabs in a window, and when you close it, it’s closed for all tabs. 
> * As a result, Chat and Skill configuration should now work properly in Edge and (mostly) in Vivaldi.
> * The lightning bolt tab that enables you to quickly open the sidepanel (aka the FAB) still doesn’t work properly in Vivaldi - you can disable it in the Settings.
> * If you encounter other issues, let me know here! 
* You can now open the Skill Picker in Chat with either  “k “ or just SPACE in the chat field. 
* When you go to the Models page, it is now more clear that you can’t enter model details until you login. Once you login, you can enter model details immediately without having to press “Add New Model”. 
* Added [Cohere Command R+](https://cohere.com/command) and Open AI GPT4o Mini as model templates. Removed OpenAI GPT 3.5
* Chat can use just the selected text on a page, but previously could not read selected text in iframes served from different domains as the main window. This has been fixed.
* For Skill Buttons - the output option “Popup” has been renamed “Floating panel” since some users thought it meant browser popup.


Notes for each release in the Chrome Web Store and Edge Add-Ons Store

Release Notes

"use strict";
const {Fragment: _Fragment, jsx: _jsx, jsxs: _jsxs} = arguments[0];
function _createMdxContent(props) {
  const _components = {
    a: "a",
    h2: "h2",
    li: "li",
    p: "p",
    span: "span",
    ul: "ul",
    ...props.components
  };
  return _jsxs(_Fragment, {
    children: [_jsxs(_components.p, {
      children: ["Ask Steve can work on your local (downloaded) files via our ", _jsx(_components.a, {
        href: "https://quickstart.asksteve.to/viewer",
        children: "File Chat"
      }), " page. This page extracts the text out of your files so that Ask Steve can use it."]
    }), "\n", _jsxs(_components.h2, {
      id: "accessing-file-chat",
      children: [_jsx(_components.a, {
        "aria-hidden": "true",
        tabIndex: "-1",
        href: "#accessing-file-chat",
        children: _jsx(_components.span, {
          className: "icon icon-link"
        })
      }), "Accessing File Chat"]
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsxs(_components.li, {
        children: ["You can open File Chat from the bottom of the Ask Steve ", _jsx(_components.a, {
          href: "/docs/context-menu",
          children: "Context Menu"
        })]
      }), "\n", _jsxs(_components.li, {
        children: ["Also from the sidebar in the extension ", _jsx(_components.a, {
          href: "https://quickstart.asksteve.to/browsers/options.html",
          children: "Settings"
        })]
      }), "\n", _jsxs(_components.li, {
        children: ["Or by clicking ", _jsx(_components.a, {
          href: "https://quickstart.asksteve.to/viewer/",
          children: "here"
        })]
      }), "\n"]
    }), "\n", _jsxs(_components.h2, {
      id: "using-file-chat",
      children: [_jsx(_components.a, {
        "aria-hidden": "true",
        tabIndex: "-1",
        href: "#using-file-chat",
        children: _jsx(_components.span, {
          className: "icon icon-link"
        })
      }), "Using File Chat"]
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsx(_components.li, {
        children: "Click where it says \"Click here to select a file\" to pick a file."
      }), "\n", _jsx(_components.li, {
        children: "Alternately you can drag and drop a file onto the purple dropzone."
      }), "\n", _jsx(_components.li, {
        children: "You can also paste an image file into this page and it will be processed."
      }), "\n", _jsx(_components.li, {
        children: "You can also paste any text into this page."
      }), "\n"]
    }), "\n", _jsxs(_components.h2, {
      id: "supported-file-types",
      children: [_jsx(_components.a, {
        "aria-hidden": "true",
        tabIndex: "-1",
        href: "#supported-file-types",
        children: _jsx(_components.span, {
          className: "icon icon-link"
        })
      }), "Supported File Types:"]
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsx(_components.li, {
        children: "pdf, docx, pptx, jpg, png, gif"
      }), "\n", _jsx(_components.li, {
        children: "For PDFs, the page tries to extract any text in the file. If it can't find any, it turns each page into an image and then OCRs it."
      }), "\n"]
    }), "\n", _jsxs(_components.h2, {
      id: "processing",
      children: [_jsx(_components.a, {
        "aria-hidden": "true",
        tabIndex: "-1",
        href: "#processing",
        children: _jsx(_components.span, {
          className: "icon icon-link"
        })
      }), "Processing"]
    }), "\n", _jsxs(_components.ul, {
      children: ["\n", _jsx(_components.li, {
        children: "All the processing is done in your browser. Your files never leave your computer."
      }), "\n"]
    })]
  });
}
function MDXContent(props = {}) {
  const {wrapper: MDXLayout} = props.components || ({});
  return MDXLayout ? _jsx(MDXLayout, {
    ...props,
    children: _jsx(_createMdxContent, {
      ...props
    })
  }) : _createMdxContent(props);
}
return {
  default: MDXContent
};


Getting Started

Working with Steve

Settings

Other

File Chat

Chat with your local (downloaded) files

Accessing File Chat

Using File Chat

Supported File Types:

Processing

Table of Contents