Practical overview
When it fits
Use Browserable when you need a developer-oriented foundation for an AI agent that performs tasks in a browser, and can configure an LLM and a remote browser provider.
Who it is for
- JavaScript developers building agents that interact with websites.
- Teams prototyping browser-based research, form-filling or information-gathering workflows.
Documented uses
- Build an agent that searches websites using criteria and returns relevant information.
- Automate browser interactions such as filling forms and clicking controls.
- Create a browser task through the JavaScript SDK and wait for its result.
Requirements & limits
The repository describes Browserable as open source, self-hostable and MIT-licensed. Its documented setup requires API keys for an LLM provider and a remote browser provider.
- The documented setup uses separate LLM and remote-browser providers; Browserable does not describe these as one bundled service.
- Manual development setup requires Docker and Docker Compose.
- The repository documents a JavaScript SDK, a REST API reference and a local dashboard for configuration.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up Browserable with Docker Compose
Start the documented development environment and configure its language-model and remote-browser provider keys.
Before you start
- Docker
- Docker Compose
- An API key for one LLM provider: Gemini, OpenAI or Claude
- An API key from one remote browser provider: Hyperbrowser or Steel
Steps
Clone the Browserable repository and change into its directory.
From the repository, change into deployment and run docker-compose -f docker-compose.dev.yml up.
Open the local admin dashboard at http://localhost:2001/dash/@admin/settings and enter an API key for one supported LLM provider.
Obtain a remote-browser provider API key and enter it in the same dashboard.
Expected outcome
The development environment is started and configured with the provider keys Browserable requires to start using it.
Guide limitations
- The instructions document development-environment setup and provider configuration, not a complete deployment or task-running procedure.
- The repository lists local services at ports 2001–2003 and other service ports; it does not describe their security or production configuration.
Create a browser task with the JavaScript SDK
Use the documented SDK pattern to submit a browser-agent task and wait for its result.
Before you start
- The browserable-js package, installable with npm or yarn
- A Browserable API key
Steps
Install browserable-js using npm or yarn.
Import Browserable from browserable-js and initialize it with an API key.
Call createTask with a task description and the BROWSER_AGENT identifier.
Use waitForRun to wait for a run result.
Expected outcome
The documented SDK pattern returns a task result that can be read from result.data.
Guide limitations
- The README example shows the task flow but does not provide a complete working code sample: its waitForRun call references taskId without showing where that value is obtained.
- The repository links to separate SDK and REST API documentation for additional implementation details.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 15 October 2026
- Original website title
- browserable/browserable · Official homepage title · retained in its source language