OpenAGI Labs

Building Computer-Use Agent Models and the open ecosystem around it.

[OpenAGI Labs @agiopen_org]

Introducing Lux, the most powerful and fastest Computer Use model.

Lux outperforms Google Gemini CUA, OpenAI Operator and Anthropic Claude on benchmark with 300 real-world tasks.

Try our developer-friendly SDK to build powerful, real-world applications.
🧵

[OpenAGI Labs @agiopen_org]

💻 Check out the Openclaw x Lux integration on GitHub: https://t.co/cHLgXFCGnw

🛠️ Build with the SDK: https://t.co/ey3Kc8jGoi

[OpenAGI Labs @agiopen_org]

OpenClaw is now a command center for a fully autonomous OS level agent.

We integrated our industry-leading Computer Use model, Lux, directly into it. Now OpenClaw can see your screen, click and type on its own.

Watch it run a complex QA test and catch bugs in seconds. 👇

[OpenAGI Labs @agiopen_org]

We are committed to building an Open Ecosystem for Computer Use.
Whether you want to use Lux out of the box or build your own custom autonomous agents, the tools are ready for you.

[OpenAGI Labs @agiopen_org]

Watch our Computer Use model visually organize a chaotic local directory directly at the OS level.

Most agents are trapped inside the browser. Lux operates natively across Mac and Windows, driving the actual desktop GUI.

In this sequence, Lux isn't running a background Python script. It uses pure visual intelligence to execute the workflow:

Perceives the UI: Visually scans the native file manager to identify file types and icons.

Synthesizes Context: Uses semantic reasoning to categorize unstructured files.

Drives the OS: Executes rapid, native GUI selections to sort them into folders.

By optimizing how the model processes UI elements, this OS-level execution is faster, cheaper, and significantly more accurate than current SOTA computer use models.

#OpenAGILabs #ComputerUse #Automation #SoftwareEngineering #AI #OS

[OpenAGI Labs @agiopen_org]

@shivsakhuja @agentmail @tryagentphone Yes, these tools are building the agent-first economy. AND now those brains need hands. The next evolution is secure Computer Use: giving agents the visual intelligence to securely take hold of the OS and drive any GUI natively. Check it out: https://t.co/tI5G1fi2pk

[OpenAGI Labs @agiopen_org]

Watch our Computer Use model visually navigate a complex Shopify dashboard to add a new product natively in the browser. Zero API integrations or CSV uploads required.

In this sequence, Lux executes a standard e-commerce operations task entirely through the GUI:

Visual Parsing: Scans the dense Shopify sidebar to locate and navigate to the 'Products' module.

Interface Interaction: Identifies and triggers the 'Add product' action within the dynamic DOM.

Data Synthesis: Maps the unstructured variables (Title, Description) into the correct rich-text input fields.

State Execution: Completes the workflow by saving the entry and visually confirming the success state.

Bypassing rigid backend integrations and driving workflows directly at the presentation layer is how automation actually scales.

#OpenAGILabs #ComputerUse #SoftwareEngineering #Automation #AI #Shopify

[OpenAGI Labs @agiopen_org]

@felixrieseberg Cross-device dispatch is exactly how agents should feel. Text the brain on mobile, let the hands work on the desktop. We're building the visual GUI execution layer for this at OpenAGI. Async delegation + native OS navigation is the future of computing.

[OpenAGI Labs @agiopen_org]

Killing the API key requirement is exactly right. The next frontier is killing the "integration" entirely. AI employees shouldn't need a connector to use an app. They should just look at the screen and drive the UI natively. That cross-app visual execution is what we're focused on with Lux.

[OpenAGI Labs @agiopen_org]

This is a great architecture: 147 specialized brains. But they still need hands. Right now, an agent can reason through a complex task, but it still relies on brittle APIs to actually execute it.
The next massive unlock is giving these models a visual execution layer so they can drive the GUI natively, just like a human operator. That’s exactly the gap we’re solving with Lux.

[OpenAGI Labs @agiopen_org]

Watch our Computer Use model visually organize an inbox and execute a multi-step workflow right in the browser. It interacts with the Gmail interface natively, zero backend integrations required.

In this sequence, Lux executes a standard operations task entirely through the UI:

Visual Navigation: Navigates the sidebar to trigger the label creation menu and types "invoices".

Native Tooling: Uses the built-in search bar to filter the inbox for financial documents.

Interface Interaction: Clicks the bulk-select checkboxes and navigates the top menu to apply the new label.

This is true zero-shot visual execution. You give a command, and the agent drives the browser.

#OpenAGILabs #ComputerUse #SoftwareEngineering #Automation #AI

[OpenAGI Labs @agiopen_org]

Watch Lux translate an unstructured chat message directly into a labeled GitHub issue. Our Computer Use model executes this cross-app GUI workflow with zero API integrations.

In this flow, Lux is doing more than just clicking coordinates. It reads the raw bug report from the chat interface, synthesizes the core technical context (iPhone 15, landscape UI overlap), and drives the browser to the repository.

It drafts the issue using structured Markdown, outlines the reproduction steps, and then physically interacts with the DOM to locate and assign the "bug" label in the sidebar before submitting.

This is how you bridge messy human communication and structured engineering trackers natively.

#OpenAGILabs #ComputerUse #SoftwareEngineering #GitHub #DevOps

[OpenAGI Labs @agiopen_org]

Social media management shouldn't require endless clicking.

What if you could trigger your entire workflow with a single command?

In this demo, Lux handles social account management autonomously. The instruction is simple: "Boost my pinned X post for 2 days at $20/day."

Watch Lux visually navigate the profile, identify the pinned post, and execute the promotion flow perfectly. No APIs. No manual setup.

Just give the command, and let the Agent execute.

#MarketingAI #SocialMediaManagement #OpenAGILabs #Automation #Tech

[OpenAGI Labs @agiopen_org]

We built Lux to handle the CRM admin work.

In this demo, our Computer Use model reads an incoming customer email about an integration failure and autonomously navigates Salesforce to log a new case. No rigid backend integrations to configure. It clicks, extracts the context, types the description, and submits the ticket exactly like a human rep does.

Stop paying top talent to act as human APIs. Let the Agent handle Salesforce.

#EnterpriseAI #Salesforce #CRM #OpenAGILabs #Automation

[OpenAGI Labs @agiopen_org]

Code-based automation is notoriously brittle. One UI update, and your entire test suite breaks.

We built Lux to solve the "Fragility Problem."

In this demo, Lux performs Autonomous Visual Validation on a desktop build. It doesn't rely on hidden code selectors. It maps the app hierarchy instantly and verifies the 'Visualizer' and 'Settings' modules exactly like a human user would.

If the backend code changes but the UI looks the same, the test still passes.

Stop maintaining broken scripts. Start deploying autonomous agents.

#EnterpriseAI #QA #DevOps #OpenAGI #Infrastructure

[OpenAGI Labs @agiopen_org]

Learn more about OpenAGI's approach to evaluating computer-use agents on our blog:

https://t.co/KPrBbolFoy

[OpenAGI Labs @agiopen_org]

Current agent evals are broken! Most teams just run the agent on tasks and report success rate. But that collapses a GUI agent's real complexity into a single binary outcome.

Here's how we think about building better evals at OpenAGI 🧵

[OpenAGI Labs @agiopen_org]

For open-ended tasks without programmatic signals, we use LLM-as-judge with per-scenario prompts that:
Describe intended behavior
Specify rubrics for correctness
Require grounding decisions in concrete trajectory evidence