Model-agnostic agent platform
OpenHands is presented as an open-source, model-agnostic platform, so teams can choose how agents run and which model providers they use.
OpenHands is an open-source, model-agnostic platform for cloud coding agents that automates engineering workflows with web, CLI, terminal UI, and SDK access.
OpenHands is an open-source platform for cloud coding agents that helps teams automate engineering work in a controlled, model-agnostic environment. The product is positioned as a way to run agents securely and transparently while keeping control over deployment, access, and data.
Across the public site, OpenHands is described as usable locally, in hosted cloud form, or in self-hosted enterprise deployments. It supports individual developers, teams, and organizations that want agents to fix bugs, review pull requests, migrate code, triage incidents, and run scheduled or event-driven automations.
OpenHands is presented as an open-source, model-agnostic platform, so teams can choose how agents run and which model providers they use.
The product can run locally, in a hosted cloud environment, or in self-hosted and private VPC deployments, giving teams several operating modes.
OpenHands supports a web GUI, terminal UI, CLI, and SDK, which lets users interact with agents in different ways and embed them into other products or internal tools.
The enterprise pages describe isolated sandboxes, repository access, PR creation, and triggered runs from GitHub, Slack, Jira, or cron workflows.
The pricing and enterprise pages list GitHub, GitLab, Bitbucket, Slack, Jira, Linear, CI/CD, SSO, RBAC, and audit-related controls as part of the platform.
Run agents that scan repositories, fix vulnerabilities, and open reviewable pull requests for engineering teams that want automated remediation with human review.
Use OpenHands to examine pull requests for quality, security, and best practices before changes are merged.
Apply the platform to legacy modernization projects, including the COBOL-to-Java migration workflow described on the homepage.
Trigger agents from GitHub, Slack, Jira, or scheduled jobs so they can investigate errors, summarize root causes, and post debugging output without requiring a laptop to stay online.
Give individual developers and teams a shared way to run automations locally or in the cloud, with organization controls when they need RBAC, audit logs, and budget guardrails.
OpenHands offers a local open-source setup, a hosted cloud option, and an enterprise deployment path. The pricing page describes an open-source local version, a SaaS Individual plan, and an Enterprise plan with self-hosted or VPC deployment options.
The source describes OpenHands as model-agnostic and says teams can use their own LLM keys or choose provider access through OpenHands plans. The enterprise page also says it works with providers such as Anthropic, OpenAI, Bedrock, and others.
OpenHands can run agents in isolated sandboxes on a VM or in the cloud, and the enterprise page says automations can be triggered from GitHub issues, Slack messages, Jira tickets, or a schedule.
The public pages show web-based GUI, terminal UI, CLI, an SDK, Git integrations, Slack, Jira, GitLab, GitHub, Bitbucket, and some enterprise controls such as RBAC, SSO, audit logs, and budget controls. The source does not provide a complete integrations catalog.