Coder

What Coder does

Coder provides self-hosted AI development infrastructure and governance for enterprises. It empowers teams to build software faster, more securely, and at scale through the collaboration of AI coding agents and human developers. Coder's mission is to make agentic AI a safe, trusted, and integral part of every software development lifecycle. The platform allows developers to use any AI model or agent within secure, high-performance self-hosted environments, while giving enterprises the governance and control to scale AI safely. It offers solutions for AI governance, centralized model access, policy-controlled environments, and full observability and auditability across all AI usage. Coder also provides standardized, self-hosted environments for developers and agents to build side-by-side with speed, consistency, and control.

Products

• Coder AgentsNative AI coding agent that runs on self-hosted infrastructure, allowing developers to delegate work through chat or API. It provides enterprise-grade security, governance, and auditability for AI coding workflows, with centralized visibility into prompts, tool calls, models, spend, and agent activity.
• Coder AI GovernanceCentralized model access and policy-controlled environments with full observability and auditability across all AI usage. It allows for centrally approving providers, models, private endpoints, and enterprise API keys, setting usage limits, and tracking costs.
• Coder WorkspacesStandardized, self-hosted environments for developers and agents to build side-by-side with speed, consistency, and control. These are open-source, ephemeral, and accessible from a browser or IDE of choice, supporting various operating systems and integrations.
• Coder CommunityA free version for hobbyists and small teams, offering unlimited workspaces and templates, unlimited members within a single organization, access to various IDEs, Coder's web UI, CLI, and API, and community support.
• Coder PremiumAn enterprise-grade offering that includes all Community features plus enhanced security, scalability, and governance. It provides ticket-based global support with SLA, multi-organization access controls, resource quotas, audit logging, unlimited Git and external auth integrations, high availability, and customizable branding.
• AI GatewayCentralizes access for coding agents like Claude Code and Codex, providing centralized authentication, user-level tracking, cost monitoring, and audit trails across all developer laptops. It enables compliant AI adoption without slowing developers down.
• Agent FirewallEnforces default-deny network policies, restricting which domains and HTTP methods agents can access. Admins define explicit allow lists in templates, and all policy decisions are logged and streamed for centralized auditing.

Key capabilities

• Self-hosted AI development infrastructure
• Governance and control for AI at scale
• Secure, high-performance self-hosted environments
• Native AI coding agent (Coder Agents)
• Centralized model access and policy control (Coder AI Governance)
• Standardized, self-hosted environments (Coder Workspaces)
• Unlimited workspaces and templates
• Support for various desktop and web-based IDEs (Cursor, VS Code, JetBrains, IntelliJ, Jupyter)
• Access to Coder's web UI, CLI, and API
• Single Sign-On (OpenID Connect)
• Ticket-based global support with SLA (Premium)
• Multi-organization access controls (Premium)
• Resource quotas per organization and user (Premium)
• Audit logging to monitor user operations (Premium)
• Unlimited Git and external auth integrations (Premium)
• High availability via multiple Coder server replicas (Premium)
• Workspace proxies for low-latency global relays (Premium)
• Automatically stop idle workspaces (Premium)
• Sync Coder groups and roles via OIDC (Premium)
• Group and user role-based access controls (Premium)
• Customize Coder's web UI with branding (Premium)
• Agent Infrastructure (MCP Server, Agent Firewall, AI Gateway)
• Support for Linux, macOS, Windows
• IDE plugins (VS Code, JetBrains)
• Coder Desktop
• Dynamic parameters
• Platform integrations (GitHub, Jfrog, Backstage)
• Notifications
• Workspace autostart/autostop
• External authentication integrations
• Prometheus metrics
• Template usage insights
• Prebuilt workspaces
• External provisioners
• Web UI notifications
• Quiet hours
• Customizable branding
• Custom roles
• Enforce workspace scheduling
• Autostop requirement
• Clean up unused workspaces
• Template permissions
• Enforce workspace updates
• Disable SSH
• Workspace command logging
• Governed workspaces with built-in auditing and observability
• Secure source code by moving it off decentralized laptops
• Optimize compute costs by right-sizing infrastructure and stopping idle environments
• Deploy anywhere (any cloud, on-premises, air-gapped)
• Integrate with existing identity providers, container registries, observability tools, and Terraform
• Any Git provider (GitHub, GitLab, Bitbucket, Azure DevOps)
• Role-based access control
• Parallel agent tasks in isolated workspaces
• AI-agnostic (supports any AI coding agent or LLM)
• Agent Firewall to define agent permissions
• Centralized visibility into prompts, tool calls, models, spend, and agent activity
• Model Governance (approve providers, models, private endpoints, API keys)
• Usage Limits & Cost Tracking for models, users, and chats
• No API Keys in Workspaces (credentials in control plane)
• Workspace Governance (restrict templates agents can use)
• User Identity on Every Action (tied to SSO)
• Standardized MCP and Skills for agent behavior
• Single control plane for agents and workspace provisioning
• Support for LLM providers: Anthropic, OpenAI, Google, Bedrock
• Trigger agents via chat UI or Coder Agents API (CI pipelines, GitHub Actions, Slack, Jira)
• Usage analytics, cost tracking, and usage limits for AI
• Self-hosted and fully inspectable agent architecture
• Authenticate every AI request and tie to named user
• Log prompts, token usage, models, and tool invocations centrally
• Contain agents behind a default-deny network
• Export audit events to SIEM

Use cases

• Empowering teams to build software faster, more securely, and at scale
• Autonomous coding with AI agents
• Scaling AI safely with governance and control
• Moving development to secure, high-performance self-hosted environments
• Building with AI agents using any AI model on secure infrastructure
• Onboarding developers and AI coding agents to new projects quickly
• Securing source code and data by centralizing development environments
• Optimizing compute costs by managing idle environments and sharing resources
• Standardizing development environments for consistency
• Providing development infrastructure as a service to internal customers
• Accelerating quantum algorithm research
• Modernizing developer onboarding for large engineering teams
• Centralizing development across classified environments
• Automating tool updates and enforcing data access controls
• Ensuring compliance with regulations like FedRAMP, GDPR, DORA
• Securely delivering consistent environments to thousands of developers
• Provisioning ephemeral workspaces as code
• Running parallel AI coding agents in isolated workspaces
• Governing agent interactions with LLM providers for auditing and cost control
• Moving AI agents into production
• Centralizing authentication, user-level tracking, cost monitoring, and audit trails for AI agents
• Restricting AI agent access with default-deny network policies
• Accelerating mission-critical innovation in the public sector
• Improving developer productivity and experience
• Reducing build times for software development

AI approach

Coder provides self-hosted AI development infrastructure and governance, allowing developers to use any AI model or agent within secure, controlled environments. It offers a native AI coding agent (Coder Agents) and an AI Gateway for centralized model access, policy enforcement, and auditability. The platform is AI-agnostic, supporting various LLM providers and allowing integration with existing CI/CD pipelines and tools.

Tech named: AI coding agents, AI models, LLM, Anthropic, OpenAI, Google, Bedrock, Claude Code, Codex

Industries served

• Software Development
• Government
• Defense
• Energy / Utilities
• Financial Services
• Technology

What it says sets it apart

• Self-hosted and deployable in any cloud, on-premises, or air-gapped environments
• Built-in governance for AI agents and developer workspaces
• Full control over permissions, audit logging, and compliance
• AI-agnostic, allowing use of any AI model or agent
• Centralized control plane for agents and workspace provisioning
• Focus on enterprise-grade security and scalability
• Empowers collaboration between AI coding agents and human developers
• Eliminates local machine constraints and inconsistent configurations
• Reduces developer onboarding time from days/weeks to hours/seconds
• Optimizes cloud costs by managing idle resources
• Provides comprehensive auditability and observability for AI usage
• Offers a native AI coding agent (Coder Agents) with enterprise features
• Supports a wide range of IDEs and operating systems
• Integrates with existing enterprise stacks (identity providers, container registries, Terraform)
• Provides an Agent Firewall for granular control over agent network access
• Ensures LLM credentials exist only in the control plane, not in workspaces
• Offers multi-organization access controls and resource quotas for large enterprises

Funding rounds we track

From the AI funding tracker — rounds as reported by the linked publications.