Cody is an AI coding assistant developed by Sourcegraph that combines large language models with deep codebase understanding to help developers write, understand, and fix code more efficiently. Designed for individual developers and enterprise engineering teams alike, Cody integrates directly into popular IDEs and leverages Sourcegraph's advanced code search index to provide context-aware assistance across entire codebases, not just the currently open file.
How It Works
Cody connects to Sourcegraph's code intelligence platform, which indexes every repository in a codebase using SCIP, a precise code indexing format. When a developer asks a question or requests a code change, Cody uses Sourcegraph's search API to pull relevant context from across the entire codebase — including APIs, symbols, usage patterns, and documentation from both local and remote repositories. This deep codebase context allows Cody to deliver more accurate and comprehensive results than AI coding assistants limited to a single file or project scope.
Key Capabilities
Cody offers a range of AI-powered features. Chat enables developers to ask natural language questions about their code, generate new code, or edit existing code, with automatic context from the open file and repository, complemented by @-mentions for adding specific files, symbols, or remote repositories. Autocomplete provides inline code suggestions as developers type. Auto-edit observes cursor movements and begins proposing contextual modifications after a single character edit. The Prompts system allows teams to create, save, and share reusable commands for automating common development workflows. Cody also includes dedicated debugging capabilities optimized to identify and fix errors, with suggestions that accelerate the debugging process.
IDE and Platform Support
Cody is available as extensions for Visual Studio Code, JetBrains IDEs, and Visual Studio. It can also be used directly from the Sourcegraph web application and through a command-line interface for terminal-based workflows. For enterprises, Cody Enterprise provides additional administrative controls, security features, and deployment options including single-tenant cloud and self-hosted environments. Cody is also compatible with other Sourcegraph products like Code Search, enabling developers to search their codebase and then jump into Cody chat with full context about any file or repository.
Pricing Model
Cody follows a freemium model. Cody Free provides access to core features including chat, autocomplete, and commands with usage limits. Cody Pro offers higher rate limits, access to more powerful models, and additional capabilities for individual developers. Cody Enterprise is designed for organizations and includes SSO, SCIM provisioning, RBAC, dedicated support, and the choice of single-tenant cloud or self-hosted deployment. Enterprise pricing starts at $16,000 and scales with team size, using a credit-based system for AI features that pools credits across the organization with no monthly expiry.
Privacy and Security
Sourcegraph collects prompts and responses to provide the service and improve the user experience, but does not use customer data to train models. Cody supports context filters that allow teams to exclude specific repositories from chat and autocomplete results, giving organizations control over what context is used. Sourcegraph is SOC 2 Type II and ISO 27001 compliant, and enterprise deployments include role-based access controls and audit logging for comprehensive security oversight.
