About
Cody is Sourcegraph's enterprise-grade AI coding assistant that goes beyond individual developer productivity to help entire engineering organizations achieve consistency, speed, and quality at scale. Trusted by 4 of the top 6 US banks, 15+ US government agencies, and 7 of the top 10 public technology companies, Cody fits seamlessly into existing workflows across all major editors and code hosts. At its core, Cody offers AI-powered code writing, fixing, and maintenance, paired with Deep Search—an agentic, natural language search capability that helps developers navigate even the most complex codebases instantly. Engineers can create and share reusable prompts to automate repetitive tasks and enforce coding best practices across their entire organization. Cody provides access to the latest large language models while maintaining strict enterprise-grade security: full data isolation, zero data retention, no model training on your code, detailed audit logs, and controlled access. It scales effortlessly across any size codebase, including massive monorepos. Additional platform capabilities include Batch Changes for large-scale cross-repository modifications, Code Insights for high-level metrics, a GraphQL and REST API, a CLI, and MCP (Model Context Protocol) support for agent workflows. Engineers using Cody report saving 5–6 hours per week and writing code 2x faster, delivering measurable ROI for organizations of all sizes.
Key Features
- Deep Search: Agentic, natural language AI search across your entire codebase, enabling developers to find relevant code and context instantly without manual navigation.
- Multi-Editor & Code Host Integrations: Works seamlessly with all major code editors (VS Code, JetBrains, etc.) and code hosts, fitting into any existing developer workflow without disruption.
- Latest LLM Access with Zero Retention: Access to the latest-generation AI models with full data isolation—your code is never retained or used to train AI models.
- Shared & Reusable Prompts: Create, share, and reuse prompts across your engineering team to automate repetitive tasks and enforce coding best practices at organizational scale.
- Batch Changes & Code Insights: Make large-scale, cross-repository code changes and track high-level code metrics and analytics to maintain quality across the entire codebase.
Use Cases
- Accelerating code development in large, complex enterprise codebases with AI-powered autocompletion and code generation
- Enforcing coding standards and best practices across engineering teams using shared, reusable prompts
- Searching and understanding unfamiliar or legacy code with natural language deep search across repositories
- Making large-scale batch code changes across multiple repositories simultaneously to reduce manual effort
- Onboarding new engineers faster by allowing them to query and explore codebases using plain-language questions
Pros
- Proven Enterprise Trust: Trusted by top US banks, government agencies, and major tech companies, demonstrating reliability and compliance at massive scale.
- Strong Security & Privacy Guarantees: Zero data retention and no model training on your code gives enterprises confidence in IP protection and regulatory compliance.
- Significant Productivity Gains: Engineers report saving 5–6 hours per week and writing code 2x faster, delivering clear and measurable ROI for development teams.
- Scales to Any Codebase Size: Handles the largest monorepos and multi-repository environments effortlessly, making it suitable for the most demanding enterprise scenarios.
Cons
- Free Plan Limitations: Recent changes to Cody Free, Pro, and Enterprise Starter plans may reduce the features and usage available on lower tiers.
- Enterprise-Oriented Feature Set: Many of the most powerful capabilities are designed for large organizations, which may feel like overkill for solo developers or very small teams.
- Complex Enterprise Onboarding: Full enterprise deployment with security controls, SSO, and integrations may require significant configuration effort and dedicated IT support.
Frequently Asked Questions
What is Cody AI?
Cody is an AI coding assistant from Sourcegraph that helps developers write, fix, and maintain code faster. It uses advanced LLMs combined with deep code graph knowledge to provide context-aware assistance across your entire codebase.
Which editors and code hosts does Cody support?
Cody works with all major code editors including VS Code and JetBrains IDEs, and integrates with all major code hosts. It also includes a CLI for terminal-based workflows.
Does Cody train AI models on my code?
No. Cody enforces full data isolation and zero data retention policies. Your proprietary code is never used to train or fine-tune AI models, making it safe for enterprise and regulated environments.
What LLMs does Cody support?
Cody provides access to the latest-generation large language models. The specific models available depend on your plan level, and Sourcegraph regularly updates support for new frontier models.
Is there a free version of Cody?
Yes, Cody offers a free plan alongside Pro and Enterprise tiers. Note that Sourcegraph has made recent changes to Free, Pro, and Enterprise Starter plans, so it's recommended to check their pricing page for the most current details.
