The easiest way to deploy browser-use in the cloud. Create always-on agents via API, or self-host with our open-source codebase.
Want a developer workflow to inspect? View the Standup Coordinator AI agent .
Simple API
Self-host or cloud
Always-on agents
Deploy browser-use agents with a few lines of code.
import requests
# Create an always-on agent
response = requests.post(
"https://api.gobii.ai/v1/persistent-agents/",
headers={"Authorization": f"Bearer {API_KEY}"},
json={
"name": "Data Collector",
"system_prompt": "You are a web research agent...",
"schedule": "every day at 9am"
}
)
agent = response.json()
print(f"Agent created: {agent['id']}")A complete platform with production-grade infrastructure, so you can focus on your product.
Real Chrome browsers running in the cloud with full rendering, JavaScript execution, and visual output. Headed or headless modes available.
Managed rotating proxies with health monitoring. Bypass bot detection automatically.
Connect agents to external tools via Model Context Protocol. Platform, org, or user-scoped.
Each agent gets its own SQLite database for structured data storage. Persisted automatically to cloud storage with compression. Query with full SQL.
Get notified when tasks complete. Push results directly to your endpoints.
Agents maintain state across runs. Cookies, auth tokens, and context preserved.
Persistent file storage for each agent. Upload, download, and share files across sessions.
The open-source standard for AI browser automation. Gobii is browser-use ready for production — deploy, scale, and manage agents without building infrastructure.
From data pipelines to customer automation.
Schedule agents to collect data from websites, APIs, and documents. Output to your database or warehouse with structured schemas.
Build agents that respond to customer inquiries via email, SMS, or chat. Integrate with your existing support stack.
Automate repetitive browser tasks: form submissions, data entry, report generation, and multi-step web workflows.
Deploy agents that monitor websites, track changes, and send alerts when specific conditions are met.
Everything you need to get started.
Gobii gives developers the API, cloud browsers, persistent sessions, files, databases, webhooks, and self-hosting path needed to run AI agents in production. Your team defines the workflow, reviews outputs, and chooses how much infrastructure to own.
Gobii's Agentic AI API lets developers create, run, and manage persistent AI agents that can browse websites, complete multi-step workflows, store state, and return structured outputs. Instead of building browser hosting, proxy handling, session persistence, file storage, scheduling, and webhook delivery from scratch, teams can call Gobii's API and deploy agents that keep working in the background. The platform is designed for browser-use style automation in production: real cloud browsers, persistent sessions, per-agent storage, structured data handling, and integration hooks. Developers can use Gobii's hosted cloud API for the fastest start or self-host the open-source platform when they need more control over infrastructure, data residency, or customization. This gives engineering teams a consistent API for agents that need to read pages, use authenticated sessions, create files, update records, and report completion without turning every workflow into a custom automation project.
Developers can build AI agents for web research, data collection, monitoring, customer communication, workflow automation, form submission, reporting, and recurring browser tasks. Common examples include agents that collect data from websites and documents, monitor pages for changes, prepare reports, respond to customer requests, update spreadsheets, or push structured results into downstream systems. Gobii works best when the workflow involves several steps across web pages, files, APIs, or tools and would otherwise require manual browsing or custom automation infrastructure. Agents can run on a schedule, preserve context across sessions, use files or an embedded SQLite database, and notify other systems through webhooks when work is complete.
browser-use is an open-source library for AI-powered browser automation. Gobii is the production platform around that kind of workflow. With browser-use alone, a developer still needs to manage browser hosting, reliability, session state, proxies, storage, scheduling, credentials, scaling, monitoring, and result delivery. Gobii is browser-use ready and provides the infrastructure layer: cloud browsers, persistent agents, file storage, embedded SQLite, webhooks, managed proxies, and deployment options. In practical terms, browser-use helps define how an AI agent operates a browser; Gobii helps teams deploy, run, and manage those agents reliably.
Gobii is different from traditional Robotic Process Automation (RPA) because it is built around AI agents that can work through open-ended browser workflows, not only fixed scripts or recorded click paths. RPA is useful when a process is stable and every screen, selector, and rule is known in advance. Gobii is better suited for workflows where the agent needs to interpret pages, collect information, compare options, adapt to changing websites, and return structured results. It is also more developer-oriented than many no-code automation tools because teams can create agents through an API, connect tools with Model Context Protocol servers, use webhooks, and choose cloud-hosted or self-hosted deployment.
Gobii agents can return structured data, files, screenshots, database records, webhook payloads, and workflow status updates depending on how the agent is configured. For data collection workflows, an agent might output normalized records, source URLs, extracted fields, notes, and confidence context. For monitoring workflows, it might send a webhook when a page changes or a condition is met. For workflow automation, it might update a file, write to its per-agent SQLite database, or hand results to another system. The important pattern is that Gobii agents are not limited to a chat transcript; they can produce reviewable, exportable outputs that engineering teams can route into existing systems.
Teams should use Gobii Cloud API when they want the fastest path to production browser agents without managing infrastructure. The cloud option is best for quickstarts, prototypes, scheduled agents, hosted browser sessions, and teams that want Gobii to handle scaling and operational overhead. Self-hosting is better when an organization needs deeper control over infrastructure, data residency, compliance requirements, network access, or internal customization. The tradeoff is operational responsibility: cloud reduces setup and maintenance, while self-hosting gives more control but requires the team to manage deployment, security, scaling, and updates.
Developer agent workflows should define what systems an agent may access, what credentials it can use, where outputs are stored, and who can review or export results. Teams should scope credentials narrowly, limit agents to approved domains and integrations, avoid unnecessary personal or sensitive data collection, and keep source links or logs that make outputs auditable. For higher-control environments, self-hosting can help with data residency, network boundaries, and internal security review. The safest pattern is to separate agent execution from human approval: let agents browse, collect, and organize work, then have people review important outputs before they trigger customer-facing, financial, or operational actions.
Create persistent agents, pass instructions, schedule work, and manage runs.
Use cloud browsers, persistent sessions, proxies, files, and per-agent storage.
Return structured data, files, screenshots, database records, and webhook payloads.
Scope credentials, approve sensitive actions, and keep important outputs auditable.
Get your API key and start running browser-use agents in minutes.