DeepTask is a desktop app for organizing work and getting AI help in one place. This guide shows you how to get started step by step. DeepTask focuses on the work itself—not the browser. You organize work into projects and tasks, chat with AI when you need help, and save important contacts. The basic flow is simple: create a project, add a task, use chat to solve problems, review the results, and repeat. When a workflow becomes repetitive, you can turn it into a reusable tool or schedule.

Step 1 - Install DeepTask

Download and install the desktop app for your system.

Download DeepTask

Then open DeepTask on your computer.

Step 2 - Complete the first-run setup

When DeepTask opens for the first time, you will go through the welcome flow.

In the setup, there is a step where DeepTask asks you to choose a model provider and enter an API key. This is needed so chat and AI features can work.

If you do not want to do this right away, you can skip it and configure it later in Settings.

Model Providers Supported In Step 2

The first-run setup currently supports these providers:

OpenRouter
Google
OpenAI
Anthropic
DeepSeek

If you are not sure which one to choose, OpenRouter is often the easiest starting point because one key can give you access to models from different providers.

How To Get Your API Key

Before you paste a key into DeepTask, make sure you are logged in to the provider you want to use. Some providers may also ask you to add billing or credits before the key works.

OpenRouter

Use OpenRouter if you want one account that can access models from several companies.

Open the key page: openrouter.ai/keys
Sign in or create an account
Create a new API key
Copy the key and paste it into DeepTask

Google

Use Google if you want to use Gemini models directly.

Open Google AI Studio: aistudio.google.com/app/apikey
Sign in with your Google account
Create an API key
Copy the key and paste it into DeepTask

Help page: Using Gemini API keys

OpenAI

Use OpenAI if you want to use OpenAI models directly.

Open the API key page: platform.openai.com/api-keys
Sign in or create an account
Create a new secret key
Copy the key and paste it into DeepTask

Anthropic

Use Anthropic if you want to use Claude models directly.

Open the key page: console.anthropic.com/settings/keys
Sign in or create an account
Create a new API key
Copy the key and paste it into DeepTask

DeepSeek

Use DeepSeek if you want to use DeepSeek models directly.

Open the key page: platform.deepseek.com/api_keys
Sign in or create an account
Create a new API key
Copy the key and paste it into DeepTask

After You Add the API Key

Once you choose a provider and save your key, DeepTask can use AI in chat and other AI-powered parts of the app.

Your key is stored on your local device.

Step 3 - Create your first project

Create one project for the thing you want to work on.

Good first examples:

"Personal Admin"
"Sales Follow-up"
"Content Plan"
"Product Launch"

Keep the first project simple and clear.

Step 4 - Add your first task

Add a small task that is easy to understand.

Good first tasks:

"Write a follow-up message for a customer"
"List the next 5 steps for this project"
"Organize contacts for this week"
"Prepare a short status update"

Small tasks make it easier to test how you want to work in DeepTask.

Step 5 - Open chat and ask for help

Now open chat and ask DeepTask to help with the task.

For example:

Help me break this task into simple steps.

Or:

Draft a first version for this task in a clear and friendly tone.

Or:

What should I do next for this project?

Step 6 - Review the result

After DeepTask helps, review the result and decide what should happen next.

You can:

mark the task as done
edit the task and try again
create a follow-up task
save the work as a repeatable tool
put the work on a schedule if it should happen again later

A Good First Workflow

If you are new to DeepTask, this is a good starting pattern:

Create one project.
Add three small tasks.
Use chat on one task.
Review the result.
Repeat for a few days before setting up tools or schedules.

This keeps the learning curve light.

Simple Tips

Write tasks in plain language.
Start small.
Use one project per clear goal.
Review results before automating anything.
Add tools only when work becomes repetitive.
Add schedules only when the task already works well.

Support & Resources

Website: deeptask.ai
Email: support@deeptask.ai
Help: Use the feedback menu inside the app to report problems or request features.