GenseeAI FAQ

General Information

What is GenseeAI?

GenseeAI is a completely free, web-based OpenClaw service that provides professional-grade AI assistance without technical overhead. It's a cloud-hosted version of OpenClaw that requires no local installation, hardware purchases, or configuration.

Is GenseeAI free?

Yes! GenseeAI is 100% free. We cover all costs including open-source LLMs and our proprietary top-ranked AI search. There are no hidden fees, no subscriptions, and no credit card required.

Who is GenseeAI for?

GenseeAI is designed for anyone who wants AI assistance without the complexity of local setup. Whether you're a researcher, software engineer, content creator, or just curious about AI assistants, GenseeAI makes it accessible to everyone.

What version is GenseeAI currently on?

The current version is OpenClaw v2026.4.2. New instances are always created on the latest version.

Setup & Getting Started

How do I get started with GenseeAI?

Getting started takes less than a minute:

1. Visit webapp.gensee.ai/ui
2. Click "Start Free — No Credit Card"
3. Sign in and you're ready to go

Everything is pre-configured with models, tools, and communication channels ready to use.

Do I need to install anything?

No installation required! GenseeAI runs entirely in the cloud. Just open your browser and start working. No hardware to buy, no software to install.

How long does setup take?

You can be up and running in under a minute. The one-click setup means everything is pre-configured and ready to use immediately.

Can I use GenseeAI on my phone?

Yes! Since GenseeAI is web-based, you can access it from any device with a browser—your phone, laptop, tablet, or desktop. Your conversations sync across devices.

How do I sign in?

1. Navigate to webapp.gensee.ai/ui in your browser.
2. You will see the GenseeAI sign-in screen.
3. Click the Sign In button.
4. Complete authentication on the login page.
5. Once signed in, you are taken to the main desktop.

How do I create a new account?

1. Go to the GenseeAI sign-in screen.
2. Below the Sign In button, click Create an account.
3. Complete the registration form.
4. After registration, sign in with your new credentials.

How do I sign out?

1. Look at the top-right corner of the screen (the Menu Bar).
2. Your name or email is displayed there.
3. Click the logout icon (an arrow pointing out of a door) immediately to the right of your name.
4. You are returned to the sign-in screen.

Features & Capabilities

What makes GenseeAI different from local OpenClaw?

Is my data secure?

Yes. Each GenseeAI instance runs in its own isolated sandbox. Your data stays private, and there's no risk to your personal files or devices. The sandboxed environment ensures separation between your AI assistant and your local system.

Can I create multiple instances?

Yes! You can create, pause, resume, snapshot, and rollback multiple instances anytime. You have full control without the complexity.

What models does GenseeAI use?

GenseeAI is pre-equipped with the best-fit open-source LLMs and our proprietary AI search. Everything is optimized out of the box—you don't need to worry about configuration.

Is GenseeAI available 24/7?

Yes! Your GenseeAI instance runs 24/7 in the cloud. You can access it anytime, even after closing your browser. You never have to worry about reachability or keeping your local machine running.

Usage & Workflow

Can I pause and resume my instances?

Absolutely. You have full flexibility to pause, resume, snapshot, and rollback your instances anytime. This gives you complete control over your workflow.

How do I access my GenseeAI after closing my browser?

Simply visit webapp.gensee.ai/ui and sign in. Your instance continues running in the cloud, and you can pick up right where you left off.

Can I continue conversations across devices?

Yes! Start a conversation on your phone and continue on your laptop without any issues. Your sessions sync seamlessly across all devices.

What tools and channels are available?

Everything comes pre-configured: models, tools, and communication channels are all ready to go. No low-level configuration needed—just start working immediately.

Sample Use Cases

The Sample Use Cases section on the welcome screen gives you ready-made guided wizards for common tasks. Click any available use case button to launch its wizard.

What use cases are currently available?

Use cases are grouped into categories on the welcome screen: All, Productivity, Sales & Outreach, Marketing, E-commerce, Finance, Research & Learn, Software Dev, and Lifestyle. You can also open the use case Wizard when you are running an instance by clicking the Wizard magic wand button next to the OpenClaw button.

How do I suggest a new use case?

Click the Suggest a Use Case button. This opens the feedback submission window for you to describe your idea to the Gensee team.

Instances

Instances are isolated OpenClaw environments. You can run up to 3 instances simultaneously.

How do I create my first instance?

From the welcome screen:

1. On the welcome screen, click New OpenClaw Instance under the “Go To Your OpenClaw Instances” section.
2. A Create Instance modal appears.
3. Choose an instance type — currently only OpenClaw is available.
4. Choose the disk source:
   • New disk — start fresh.
   • From existing save point — restore from a previously saved state.
5. Select the version (image) to use from the dropdown list (the latest version is pre-selected).
6. Edit the instance name if desired (a default name with the current date/time is provided).
7. Click Create Instance.
8. Wait for the instance to start (a spinner is shown during initialization).

From the dock:

1. Click the + (plus) icon in the dock at the bottom of the screen.
2. Follow steps 2–8 above.

How do I upgrade an existing instance to the latest version?

New instances always use the latest OpenClaw version (currently v2026.4.2). To upgrade an older instance:

1. Pause the instance (this creates an automatic save point).
2. Resume it — when the instance restarts, it automatically picks up the latest version.

All your files, settings, and installed skills are preserved across the restart. You can verify the version in Settings → Debug Info.

How do I resume a paused instance?

From the main window:

1. Select the paused instance in the dock or via Instances Manager.
2. The main window shows a "This instance is paused. Resume it to continue." message with a green Resume Instance button.
3. Click Resume Instance.

From the dock:

1. Make sure the paused instance is the active one.
2. Click the Play/Pause button (▶/⏸) in the Instance Controls section of the dock.

How do I pause a running instance?

Pausing stops the sandbox and automatically creates an "automatic" save point so you can resume exactly where you left off.

1. Make sure the instance you want to pause is the active one.
2. In the bottom dock, click the Play/Pause button (⏸) in the Instance Controls group.
3. A progress overlay appears while the instance is pausing.
4. A toast notification confirms: "Instance paused successfully."

How do I switch between instances?

1. Click the Layers icon in the dock — this opens Instances Manager.
2. A full-screen overlay shows all your instances as cards with their names, status, and creation dates.
3. Click any card to switch to that instance.
4. Instances Manager closes and the selected instance becomes active.

Alternatively, click Open Existing Instances on the welcome screen to open Instances Manager.

Option A — from the main window (when the instance is paused):

1. Click the small pencil icon next to the instance name in the main window.
2. Type the new name in the text field that appears.
3. Press Enter or click elsewhere to confirm.

Option B — from Instances Manager:

1. Open Instances Manager (Layers icon in the dock).
2. Hover over an instance card — a pencil icon appears in the top-right corner.
3. Click the pencil icon, type the new name, and press Enter or click away to confirm.

How do I clone an instance?

Cloning creates a new, independent instance that starts from the same state as the current running instance — useful for branching your work.

1. Make sure the instance you want to clone is running.
2. In the dock, click the Instance Controls button (sliders icon).
3. Click Clone.
4. In the Clone Instance modal:
   • Edit the name for the new clone (defaults to "Clone of [original name]").
   • Choose the software version for the clone.
5. Click Clone Instance.
6. A toast message appears: "Cloning… This may take a few minutes." The cloning runs in the background.
7. Use Instances Manager to switch to the new cloned instance when complete.

How do I delete an instance?

1. Open Instances Manager (Layers icon in the dock).
2. Hover over the instance card you want to delete — a red trash icon appears.
3. Click the trash icon.
4. A confirmation dialog asks you to confirm. Click Delete.

What happens when I reach the instance limit (3 running)?

When you try to create or resume a 4th running instance, a modal titled "Instance Limit Reached" appears, showing your 3 currently running instances. For each, you can:

• Open — Navigate to that existing instance.
• Pause — Pause that instance to free up a slot; the Create Instance modal opens automatically afterward.
• Delete — Permanently remove that instance; the Create Instance modal opens automatically afterward.

Views & Windows

The dock at the bottom contains buttons to open different views and windows for the active instance.

How do I open the OpenClaw dashboard?

Click the OpenClaw logo icon in the dock (leftmost icon). This brings the main OpenClaw dashboard into focus and hides any other floating windows in front of it.

How do I open the Browser view?

Click the Globe icon in the dock. A floating window opens showing a browser running inside your instance. You can navigate URLs, interact with web pages, and watch OpenClaw perform browser-based tasks in real time.

How do I open the File Manager?

Click the Folder icon in the dock. A floating window opens with the File Manager for your instance's workspace.

How do I open Settings?

Click the Settings (⚙) gear icon in the dock. A floating window opens with sections for Welcome, Channels, Models, Connectors, OpenClaw Control, and Debug Info.

How do I open Instances Manager?

Click the Layers icon (⧉) in the dock to open Instances Manager. This shows all your instances as cards — click any card to switch to it. You can also rename or delete instances from here.

How do I close or hide a floating window?

Click the red close button (a small red dot/circle) in the top-left corner of any floating window to close it. You can also click the corresponding dock icon again when the window is already in the foreground — this hides it behind the dashboard.

Save Points & Reverting

Save points (snapshots) capture the complete state of your instance so you can roll back to any earlier point at any time.

How do I create a save point?

Method A — Quick save from the dock:

1. In the dock, click the Instance Controls button (sliders icon).
2. Click the Camera (snapshot) icon.
3. A small input popup appears with a pre-filled name like "Snapshot taken at [date/time]".
4. Edit the name if needed, then press Enter or click the checkmark button.
5. A success toast confirms the save point was created.

Method B — Save Point modal:

1. In the dock, click the Instance Controls button.
2. Click Save Point.
3. The Create Save Point modal opens with a pre-filled default name.
4. Edit the name if desired.
5. Click Create Save Point (or press Enter).
6. A toast notification confirms: "Save point '[name]' created."

How do I revert to a previous save point?

1. In the dock, click the Instance Controls button.
2. Click Revert (clock/history icon).
3. The Revert to Snapshot panel slides open, showing a chronological timeline of all your save points.
4. Find the save point you want to restore. Each card shows the name and creation time. Auto-created save points are labeled "Auto."
5. Click the Revert button (↺) on the desired save point card.
6. A progress overlay appears. If the instance is running, it is paused first, then the snapshot is restored and the instance resumes from that state.
7. A toast confirms: "Reverted to '[name]'."

How do I rename a save point?

1. Open the Revert Panel (Instance Controls → Revert).
2. Hover over the save point card you want to rename — a small pencil icon appears next to the name.
3. Click the pencil icon.
4. An inline text field appears. Type the new name.
5. Press Enter or click the checkmark to confirm.

How do I delete a save point?

1. Open the Revert Panel (Instance Controls → Revert).
2. On the save point card you want to delete, click the trash icon.
3. A confirmation dialog appears: "Are you sure you want to delete [name]? This action cannot be undone."
4. Click Delete to confirm.

Autopausing

Does GenseeAI automatically pause inactive instances?

We're seeing a large and growing number of enthusiastic users actively building on GenseeAI.

To ensure every user actively using GenseeAI receives enough computing resources, we automatically pause instances that have been inactive for more than 48 hours.

A few important notes:

• You can resume your instance at any time with a single click of the Resume button.
• Your files, chat history, and configuration all remain safe and intact during the pause.
• Want to keep your instance running continuously? Simply use GenseeAI regularly! Auto-pause thresholds depend on your plan — see Pricing for details.

What counts as active?

Any of the following count as active use, whether they are done by you (human) or triggered by a scheduled cron job (agent):

• Instance operations — logging in or out, creating an instance, taking a snapshot, resuming or pausing an instance, and similar actions.
• Browser operations — any web browsing performed by the AI inside your instance. We only count operations, but we cannot see the actual operations.
• File operations — any file activity by you or the AI, such as uploading, downloading, copying, deleting, or moving files. We only count operations, but we cannot see the actual operations or the file content

Just one action by you or your cron jobs in any of these categories within a 48-hour period is enough to keep your instance active. We'll continuously review this policy and we'd love to hear your thoughts — share your feedback and help us make it better.

Rescue Mode

My OpenClaw instance is stuck — what can I do?

If your instance becomes unresponsive or fails to start, here are your options:

• If you previously created a save point, you can revert to it. See How do I revert to a previous save point?
• If you recently changed your OpenClaw config, you can reset or edit it without starting over. See How do I reset or edit my OpenClaw config while my instance is stuck?
• If you're comfortable starting fresh, you can create a new OpenClaw instance.
• If none of the above options work, please contact our support and we'll do our best to help.

How do I reset or edit my OpenClaw config while my instance is stuck?

To enter rescue mode:

1. Visit webapp.gensee.ai/ui/?sos=1 in your browser. Make sure the URL contains ?sos=1.
2. Click your stuck instance. It must be in paused state. If it isn't, open Instance Controls and pause it first.
3. A Rescue button will appear — click it to boot into rescue mode.

Once in rescue mode, you can:

• Download your current OpenClaw config as a backup.
• Upload a config file to overwrite the existing one.
• Reset the config to its original default state. This will usually fix the unresponsive instance problem.

To exit rescue mode, pause the instance (Instance Controls → Pause), then resume it normally using the resume button.

Security & Privacy

Is GenseeAI safe for my personal files?

Yes. The isolated sandbox approach means your AI assistant can help you without any risk to your personal files or devices. Your local data stays separate from the cloud environment.

Who can access my data?

Your data remains private within your isolated sandbox. GenseeAI provides the infrastructure, but your conversations and work are yours.

What happens if I have security concerns?

The sandboxed architecture is designed specifically for security-conscious users. Each instance is completely isolated, so you can use AI assistance without worrying about exposing your local system.

Pricing & Billing

What plans does GenseeAI offer?

GenseeAI has three plans: Free ($0/month), Premium ($29/month or $290/year), and Premium+ ($39/month or $390/year). The Free plan includes 1 instance, 10 GB storage, and 20M tokens/month of Gensee inference. Premium and Premium+ unlock more instances, more storage, longer auto-pause windows, terminal access, and higher token allocations. See our Pricing page for full details.

Do I need a credit card to sign up?

No. The Free plan requires no credit card and no payment. You only need to provide payment details if you choose to upgrade to Premium or Premium+.

What are Gensee inference tokens?

Gensee includes built-in access to a frontier AI model (Qwen 3.5-397B) and #1 ranked AI web search. Your monthly token allocation covers both model inference and web search queries. If you bring your own API key from OpenAI, Anthropic, Google, or other providers, those calls don't count against your Gensee token limit.

Can I cancel anytime?

Yes. Cancel from your account settings at any time. Your instances and data are preserved — you'll revert to Free plan limits at the end of your billing period. No cancellation fees.

Communication Channels

What are channels?

Channels are communication pathways that let you interact with your OpenClaw assistant from your favorite messaging apps. Instead of always opening the web interface, you can talk to your AI assistant directly from WhatsApp, Telegram, or other supported platforms.

Which channels are currently supported?

We currently support:

• WhatsApp
• Telegram
• Slack

All channels can be set up by following the step-by-step wizard in Settings → Channels.

How do I connect WhatsApp to my instance?

WhatsApp setup has two steps. Open Settings (gear icon in the dock) → Channels → WhatsApp.

Step 1: Basic Information

1. Choose the type of number you will use:
   • Dedicated Number — a separate number used only for this bot.
   • Personal Number — your existing personal WhatsApp number.
2. Enter your phone number including the country code (e.g. +1 234 567 8900).
3. Click Save & Continue.

Step 2: Link WhatsApp (QR Code)

1. Wait for the QR Code to appear (this can take up to a minute).
2. On your phone, open the WhatsApp app.
3. Tap the camera icon at the top of the chats screen (or go to Settings → Linked Devices).
4. Scan the QR code shown on screen. You may need to scan it twice.
5. WhatsApp will confirm the device is linked.

After I set up WhatsApp, what if it is not taking effect? I sent a WhatsApp message to myself and didn't see the OpenClaw response.

You may need to reload the OpenClaw config. See How do I reload the OpenClaw config? for how to do it.

How do I connect Telegram to my instance?

Telegram setup has three steps. Open Settings → Channels → Telegram.

Step 1: Create a Bot

1. Open the Telegram app and start a chat with @BotFather.
2. Send the command /newbot and follow the prompts to set a display name and username.
3. @BotFather will reply with a Bot Token (e.g. 123456:ABC-DEF1234...). Copy it.
4. Click I have my token → to proceed.

Step 2: Enter the Bot Token

1. Paste your bot token into the Bot Token field.
2. Click Save & Continue.

Step 3: Approve Pairing

1. Open Telegram and start a chat with your newly created bot.
2. Send /start to the bot. It will reply with a pairing code.
3. Enter or confirm the pairing code as instructed on screen.

How do I use the OpenClaw Gateway UI channel?

Open Settings → Channels → OpenClaw Gateway UI. This option lets you interact with OpenClaw directly through the built-in gateway interface without requiring a third-party messaging app. Follow the on-screen instructions to access the gateway URL.

How do I connect Slack to my instance?

Slack setup has four steps. Open Settings (gear icon in the dock) → Channels → Slack.

Step 1: Create a Slack App

1. Click Open Slack App Creation to open the Slack API site in a new browser tab.
2. Select "From a manifest" and choose the workspace you want to connect.
3. Switch to YAML format and paste the manifest shown in the wizard (click Copy to grab it).
4. Review the permissions and click Create.
5. On the app's Settings page, go to Basic Information → App-Level Tokens. Click Generate Token, name it (e.g. "socket"), add the connections:write scope, and click Generate. Copy the App Token (starts with xapp-).
6. Go to OAuth & Permissions in the sidebar, click Install to Workspace and approve. Copy the Bot User OAuth Token (starts with xoxb-).

Step 2: Enter Your Tokens

1. Paste the App Token (xapp-...) and the Bot User OAuth Token (xoxb-...) into the corresponding fields.
2. Click Save & Continue.

Step 3: Save Configuration

The wizard saves your tokens and channel configuration automatically. Wait for the confirmation message before continuing.

Step 4: Verify Connection

The wizard runs a quick status check to confirm the Slack channel is configured. If the gateway hasn't restarted yet, this may show a timeout — that's normal. You can restart the instance (pause → resume) to apply the new config.

After Setup

1. In Slack, click Apps in the left sidebar, search for your bot name, and click Add.
2. In each channel you want the bot to read and write messages, type /invite @YourBotName.
3. Send the bot a direct message or @mention it in an invited channel to start chatting.

Will you add more channel support?

Yes! We're actively working on adding support for:

• Discord
• iMessage
• DingTalk (Dingding)
• Feishu (Lark)

Stay tuned for updates as we roll out these integrations.

Browser-Based Tasks

What are browser-based tasks?

Unlike other cloud-based OpenClaw hosting providers, GenseeAI provides full browser automation support. When you need OpenClaw to browse the web, perform research, fill out forms, or interact with websites, it will automatically open web pages in our secure Browser window.

How do browser-based tasks work?

When you request a task that requires web interaction:

1. OpenClaw will automatically open the necessary web pages in our Browser window
2. The assistant navigates, searches, and interacts with sites on your behalf
3. For tasks requiring login or user input, you can interact directly with the Browser window (e.g., entering login credentials)
4. Once complete, OpenClaw summarizes the results for you

Is my login information safe?

Absolutely. Any credentials you enter in the Browser window are not saved and are completely private. The session is isolated, and your login information is never stored or shared.

What kinds of tasks can browser automation handle?

• Research and information gathering
• Form filling and submissions
• Website testing and verification
• Multi-step workflows across websites
• Capturing content from web pages
• Automating repetitive web tasks

File Management

The File Manager lets you browse, edit, upload, and download files in your instance's workspace. Open it by clicking the Folder icon in the dock.

How do I navigate folders?

• Double-click a folder to open it.
• Use the breadcrumb trail at the top to navigate back up the hierarchy.
• Click Workspace in the left sidebar to jump to the root.
• Use the ← / → arrows in the toolbar to move through browsing history.

How do I upload a file?

1. Navigate to the folder where you want to upload.
2. Click the Upload button (↑ arrow icon) in the toolbar.
3. A system file picker opens — select one or more files.
4. The files are uploaded and the listing refreshes automatically.

How do I download a file or folder?

Single item: hover over a file or folder and click the download icon. Folders are downloaded as ZIP.

Multiple items: check several items using the checkboxes on the left, then click Download in the action bar. A dialog lets you name the ZIP file before downloading.

How do I create a new folder?

1. Navigate to the parent directory.
2. Click the New Folder button in the toolbar.
3. An inline input bar appears — type the folder name and press Enter or click Create.

How do I rename a file or folder?

1. Hover over the file or folder — a pencil icon appears.
2. Click the pencil icon, edit the name, then press Enter to confirm or Escape to cancel.

How do I delete a file or folder?

1. Hover over the file or folder — a trash icon appears.
2. Click the trash icon and confirm the dialog.

How do I edit a file?

1. Click a text, code, or Markdown file to open its preview.
2. In the preview pane, click the Edit (pencil) button.
3. A text area opens with the file's content. Make your edits.
4. Click Save to write changes back, or X to discard.

What file types can I preview?

• Images (PNG, JPG, GIF, SVG, etc.) — rendered inline.
• PDFs — rendered inline in an embed viewer.
• Markdown files (.md, .mdx) — rendered as formatted HTML.
• Code files (.py, .js, .ts, etc.) — displayed with monospace formatting.
• Text files (.txt, .log, .csv) — displayed as plain text.

How do I sort and multi-select files?

Click any column header (Name, Type, Size, Modified) to sort by that column. Click again to reverse the order. Folders are always pinned above files.

To select multiple items, click the checkbox on the left side of each row. Hold Shift to select a range, or ⌘ / Ctrl-click to toggle individual items. When items are selected, an action bar appears with Copy to…, Move…, Download, and Delete.

Cloud Storage

The File Manager supports Google Drive, Dropbox, and OneDrive. Connect a provider in the left sidebar under Cloud Storage.

• Browse cloud files just like workspace files.
• Click the copy icon on any file to open the destination picker — choose where to copy it (workspace or another cloud provider) and an optional destination folder.
• Multi-select works across sources: select several files and click Copy to… to copy them all in one go. Folders are recursively copied.
• Move is supported within the same source (workspace or a single cloud provider).

How do I connect Google Drive, Dropbox, or OneDrive?

1. Open the File Manager (folder icon in the dock).
2. In the left sidebar, expand Cloud Storage.
3. Click the link icon next to the provider you want to connect (Google Drive, Dropbox, or OneDrive).
4. An OAuth popup opens — sign in with your account and grant access.
5. Once connected, click the provider name in the sidebar to browse its files.
6. Use the copy icon on any file to copy it between your workspace and the cloud provider.

To disconnect a provider, hover over its name in the sidebar and click the unlink icon.

Models & AI

What models does GenseeAI use?

GenseeAI uses a combination of open-source models and dynamically chooses the best one to serve each request. All open-source models are served by Gensee and offered completely free to you.

Can I bring my own models?

Yes! You can bring your own models by simply adding your API keys in Settings → Models. This gives you flexibility to use specific models or providers if you prefer.

How do I set the default AI model for my instance?

1. Open Settings (gear icon in the dock) → Models.
2. At the top, click the Default Agent Model dropdown.
3. Select your preferred model. Models are grouped by provider:
   • Gensee Hosted Models — always available, no key required.
   • OpenAI, Anthropic, Google Gemini — require API key setup (see below).
4. The setting saves automatically.

How do I use the free Gensee-provided models?

Free models are included with GenseeAI and require no setup. In the Default Agent Model dropdown, select from the Gensee Hosted Models group — no API key or extra cost needed.

How do I set up an OpenAI API key?

1. Go to Settings → Models → OpenAI Models.
2. Click the toggle to enable OpenAI and enter your API key (starts with sk-).
3. Click Continue, then confirm the restart dialog by clicking Restart Now (~1–2 min).
4. After restart, set your Default Agent Model to an OpenAI model.

How do I set up an Anthropic (Claude) API key?

1. Go to Settings → Models → Anthropic Models.
2. Click the toggle to enable Anthropic and enter your API key (starts with sk-ant-).
3. Click Continue, then Restart Now. After restart, set your Default Agent Model to a Claude model.

How do I set up a Gemini API key?

1. Go to Settings → Models → Gemini Models.
2. Click the toggle to enable Gemini and enter your API key (starts with AIza).
3. Click Continue, then Restart Now. After restart, set your Default Agent Model to a Gemini model.

How do I disable an AI provider and remove its API key?

1. Go to Settings → Models and expand the provider you want to disable.
2. Click the toggle to disable it.
3. Confirm the Disable & Restart dialog. After the restart, the key is removed from your instance.

How does the model selection work?

OpenClaw intelligently routes each request to the most appropriate model based on the task requirements, ensuring optimal performance while keeping everything free and seamless.

Connectors & Integrations

What are Connectors?

Connectors are integrations that extend what OpenClaw can do. For example, the Google Workspace connector lets OpenClaw read and send Gmail, manage your Google Calendar, read and edit Google Drive files, and more. Connectors are configured through Settings → Connectors, and you can add more integrations over time as they become available.

How do I install the Google Workspace (GWS) Connector?

The GWS connector requires you to set up a Google Cloud project with OAuth credentials so OpenClaw can access your Google account on your behalf. The setup takes about 10–15 minutes and is done entirely through Settings → Connectors → Google Workspace Connection.

The setup wizard walks you through 8 steps:

Step 1 — Create a Google Cloud Project

1. Go to console.cloud.google.com and sign in with the Google account you want to connect to OpenClaw.
2. Click the project dropdown at the top of the page (next to the Google Cloud logo) and click New Project.
3. Enter a project name (e.g. "OpenClaw GWS") and click Create.
4. Make sure your new project is selected in the top dropdown before proceeding.

Step 2 — Enable the APIs You Need

1. In the Google Cloud Console, go to APIs & Services → Library.
2. Search for and enable each API you want OpenClaw to access. Common ones include:
   • Gmail API — for reading and sending email
   • Google Calendar API — for reading and managing calendar events
   • Google Drive API — for reading and editing Drive files
   • Google Docs API — for working with Docs
   • Google Sheets API — for reading and writing Sheets
   • Google Slides API — for working with Slides
   • Google Tasks API — for managing Tasks
   • Google Forms API — for reading Form responses
   • People API — for reading contacts
3. Only enable the APIs for the skills you intend to install — OpenClaw will only work with the skills whose APIs are enabled here.

Step 3 — Configure the OAuth Consent Screen

1. Go to APIs & Services → OAuth consent screen (also called "Branding" in newer Google Cloud UI).
2. When asked for the User Type:
   • Choose Internal if your Google account belongs to a Google Workspace organisation — this is the easiest option as it skips the review process.
   • Choose External if you are using a personal Gmail account. Your app will be placed in "Testing" mode, which is fine for personal use.
3. Fill in the required fields: App name, User support email, and Developer contact email. The app name and logo are what you see on Google's OAuth permission screen — you can use anything you like (e.g. "My OpenClaw").
4. Click Save and Continue through the Scopes and Summary screens (no changes needed there).

Step 4 — Add Test Users (External accounts only)

If you chose External in Step 3, your OAuth app will be in Testing mode. Google restricts sign-in to only the Google accounts you explicitly add as test users — anyone else will see an "access blocked" error. You must add the Google account you intend to connect to OpenClaw here.

1. In the OAuth consent screen page, go to the Test users section.
2. Click + Add Users and enter the email address of the Google account you will connect to OpenClaw.
3. Click Save.

If you chose Internal in Step 3, skip this step.

Step 5 — Create an OAuth Client ID

1. Go to APIs & Services → Credentials.
2. Click + Create Credentials and choose OAuth client ID.
3. For Application type, select Desktop app.
4. Give it a name (e.g. "OpenClaw Desktop") and click Create.
5. A dialog appears showing your Client ID and Client Secret. Click Download JSON to download the credentials file.

Step 6 — Store the Credentials in OpenClaw

1. Open the downloaded JSON file in a text editor and copy its entire contents.
2. Back in the Connectors setup wizard in OpenClaw, paste the JSON into the credentials field.
3. Click Save Credentials.

Step 7 — Authorize Your Google Account

1. OpenClaw will display a Google authorization URL. Click the link to open it in your browser.
2. Sign in with the Google account you want to connect (must be one of the test users if you chose External).
3. Review the permissions and click Allow.
4. Google will redirect you to a page — copy the full URL from your browser's address bar and paste it back into the OpenClaw field, then click Submit.

Step 8 — Select and Install Skills

1. OpenClaw loads the available Google Workspace skill categories (Gmail, Calendar, Drive, Docs, Sheets, etc.).
2. Select the skill categories you want to install. Only skills whose corresponding Google APIs you enabled in Step 2 will work — others will appear greyed out or fail silently.
3. Skills you have already installed appear greyed out with a green checkmark — you don't need to re-install them.
4. Click Install Selected Skills.
5. Click Done to close the Settings window. OpenClaw can now use the installed skills.

Why does the OAuth app show a warning saying it's unverified?

This is expected. Because you created the Google Cloud project yourself and it hasn't gone through Google's app verification process (which is intended for published apps), Google shows a warning screen. Since you created the app yourself and you are the only user, this is safe to proceed. Click Advanced then Go to [App Name] (unsafe) to continue.

Which skills are available?

Currently available Google Workspace skill categories include: Gmail, Calendar, Drive, Docs, Sheets, Slides, Forms, Tasks, and Contacts. More skills and integrations will be added to the platform over time.

Can I connect multiple Google accounts?

Each instance supports one connected Google account at a time. If you need to switch accounts, you would need to re-run the authorization flow (Steps 7–8) with the new account.

Is it safe to connect my Google account?

OpenClaw is an AI and can make mistakes — including irreversible ones, like accidentally deleting emails or calendar events. Be careful about what you ask it to do, double-check destructive actions, and start with low-stakes tasks on an account you're comfortable experimenting with. Only enable the APIs you actually need.

Can I install skills by asking OpenClaw directly?

Yes. You can ask OpenClaw in the chat to install skills for you — just describe what integration you want to add. OpenClaw can look up and install skills from its skill registry on your behalf without going through the Settings wizard.

How do I set up GitHub?

Go to Settings → Connectors → GitHub Connection Setup. The wizard walks you through 4 steps:

1. Install — the gh CLI is downloaded automatically.
2. Authenticate — a device code is generated; type it in the GitHub authorization page that opens in the built-in browser.
3. Verify — confirms gh auth status succeeds and adds gh to your PATH.
4. Get Started — try sample prompts like "Show my GitHub repos" or "Check my PR status".

Once connected, use cases like Website Building and Documentation Generator can clone, create, and push to repos directly.

Are more connectors coming to the Settings wizard?

Yes — we are actively building more guided connector setup flows in the Settings wizard (Settings → Connectors). Future releases will add step-by-step wizards for additional integrations so you can set them up without any manual configuration.

Settings — OpenClaw Control

Open Settings by clicking the gear icon in the dock, then select OpenClaw Control in the left sidebar.

What is OpenClaw Control?

OpenClaw Control is a panel in Settings that lets you manage your OpenClaw configuration directly. It provides three actions: downloading your current config, uploading a modified config, and reloading OpenClaw to apply changes.

How do I download my config?

Click Download Config to save a copy of your current OpenClaw configuration as a JSON file (openclaw-config.json). Sensitive values like API keys and tokens are automatically redacted in the download, so the file is safe to share for debugging purposes.

How do I upload a config?

Click Upload Config and select a .json file from your computer. You'll see a confirmation dialog before anything changes — uploading will overwrite your existing configuration. We strongly recommend downloading a backup first.

After uploading, OpenClaw reloads automatically. Your connection will be briefly disrupted (up to a few minutes) while it restarts.

How do I reload OpenClaw?

Click Reload OpenClaw when you've made configuration changes that require a restart — for example, after connecting a new channel or changing model settings. OpenClaw will restart and you'll see a status indicator while it comes back online. This usually takes under a minute but can take up to 5 minutes.

What if OpenClaw doesn't come back after a reload or upload?

First, try refreshing the page to get the latest status. If that doesn't help, use the Pause and Resume buttons on your instance to restart it. If the instance is still stuck, see the Rescue Mode section for recovery steps.

Context and Image Limit

What is the context limit and what should I do when I reach it?

OpenClaw has a limited context window — the amount of conversation and information it can hold in memory during a single session. You can also share up to 8 images per session. When you approach the limit, OpenClaw will display a warning in the chat.

To start fresh without losing any files or settings, you have two options:

• Type /new in the OpenClaw chat (or send it via WhatsApp or Telegram).
• Click the New Session button in the OpenClaw dashboard.

Starting a new session only clears the conversation context — your files, installed skills, model settings, and all configurations are completely unaffected.

Technical Details & Troubleshooting

Where is GenseeAI hosted?

GenseeAI runs in the cloud on GenseeAI's infrastructure. This means you don't need to worry about server maintenance, updates, or uptime.

How large is the user disk?

Disk size depends on your plan: 10 GB on the Free plan, 50 GB on Premium and Premium+. See Pricing for details.

What happens if GenseeAI's servers go down?

As a cloud service, GenseeAI maintains infrastructure with high availability. Any downtime or maintenance would be communicated through official channels.

Can I export my conversations?

This feature may be available in future updates. Check the platform for the latest capabilities.

How do I check what is happening inside my instance?

Open the Debug Info panel in Settings (gear icon → Debug Info) to see:

• Sandbox ID
• Container ID
• Storage ID
• Image Name
• Created and Expiry timestamps with a live countdown

How do I find my instance's expiry time?

1. Open Settings (gear icon in the dock).
2. Select Debug Info in the left sidebar.
3. The Expires At row shows the expiry timestamp and a live countdown (green if plenty of time remains, amber if under 1 hour, red if expired).

How do I reload the OpenClaw config?

Go to Settings → OpenClaw Control and click Reload OpenClaw. See the OpenClaw Control section for details.

What do I do if my instance is stuck initializing?

1. Wait up to 2–3 minutes for the instance to fully start. The page auto-refreshes every 30 seconds.
2. If the instance does not start, try pausing it and resuming again.
3. If the problem persists, use the Feedback button in the dock to report the issue, or chat with the Gensee Agent using the chat widget at the bottom of the page — the agent can help resolve most issues immediately.

How do I send feedback or report a bug?

1. In the dock, click the Feedback button (speech bubble icon).
2. A Feedback modal opens — describe your issue or suggestion and submit the form.

You can also use the Gensee Agent chat widget at the bottom of the page for real-time help or to escalate to a human support agent.

How do I reset an instance to a clean state?

Revert to the earliest save point to get as close to the original state as possible:

1. Open the Revert Panel (Instance Controls → Revert).
2. Scroll to the oldest save point in the timeline.
3. Click the Revert (↺) button on that save point.

If no save points exist, create a new instance from scratch via Create Instance → New Disk.

Support & Contact

How do I get help?

For support, visit gensee.ai/contact.

Where can I learn more?

• Blog: gensee.ai/blog
• About: gensee.ai/about
• Privacy Policy: gensee.ai/privacy
• Terms of Service: gensee.ai/terms

Where is GenseeAI located?

Gensee, Inc.
9510 Innovation Ln, Room 431
La Jolla, CA 92093

Plans & Limits

What's included in the Free plan?

The Free plan gives you 1 running instance with 1 vCPU, 4 GB RAM, 10 GB storage, and 20M tokens/month of Gensee inference (or bring your own API key for unlimited). Instances auto-pause after 30 minutes of inactivity. All core features — file manager, browser, checkpoints, rollback, WhatsApp/Telegram/Slack channels — are included.

What do Premium and Premium+ add?

Both paid plans give you up to 3 concurrent instances, 4 vCPU, 6 GB RAM, 50 GB storage, terminal (command line) access, and auto-pause extended to 7 days. Premium includes 20M priority tokens/month; Premium+ includes unlimited priority tokens (fair use) and technical support by human experts. See Pricing for full details.

What does "bring your own API key" mean?

On any plan, you can use your own API key from OpenAI, Anthropic, Google, or other providers. Those calls go directly to the provider — they don't count against your Gensee token limit, and there's no markup. This is useful if you want unlimited inference on the Free plan or prefer a specific model.

How do I provide feedback?

Reach out through our contact page with any suggestions, bug reports, or feature requests.