Sourcegraph Cody is an AI coding assistant built on top of Sourcegraph's powerful code search and intelligence platform, designed to provide context-aware AI assistance that understands entire codebases rather than just individual files. Sourcegraph was founded in 2013 by Quinn Slack and Beyang Liu, and is headquartered in San Francisco, California. The company originally established itself as the leading code search and navigation platform, and Cody extends those capabilities with AI that has deep, comprehensive understanding of code repositories. Cody's primary differentiator is its ability to retrieve relevant context from across an entire codebase using Sourcegraph's code graph and search infrastructure. When a developer asks Cody a question or requests code generation, it automatically searches the full repository to find the most relevant code, documentation, and architectural patterns, providing responses that are grounded in the actual codebase rather than generic training data. This context-awareness makes Cody particularly effective for large, complex codebases where understanding cross-file dependencies and organizational conventions is critical. Cody provides autocomplete suggestions as developers type, a chat interface for asking questions about code and requesting explanations or modifications, and inline editing commands for refactoring and generating code. It supports multiple LLM backends including Claude, GPT-4, Gemini, and Mixtral, allowing organizations to choose their preferred model or use different models for different tasks. Cody is available as extensions for VS Code, JetBrains IDEs, Neovim, and the Sourcegraph web interface. Sourcegraph offers Cody Free with limited usage, Cody Pro at $9 per month for individual developers with higher limits, and Cody Enterprise with custom pricing that includes full Sourcegraph code search, multi-repository context, custom model selection, and enterprise security features including SSO, audit logs, and self-hosted deployment options.

AI 代码助手

Sourcegraph Cody 是一个以其深厚代码库意识而著称的 AI 编码助手，它利用 Sourcegraph 的代码搜索基础设施从整个存储库中检索相关上下文。它提供自动补全、聊天和内联编辑，其响应基于实际代码库模式而不是通用建议，特别适合大型、复杂的代码库。

AI 代码生成工具

Cody 生成的代码在上下文中适合其所处理的特定代码库，使用 Sourcegraph 的代码图来理解现有的模式、约定和依赖关系。开发人员可以描述所需的功能，并接收遵循其项目的架构模式和编码标准的生成代码。

AI 代码审查

Sourcegraph Cody 通过解释复杂代码更改、识别潜在问题以及提供有关修改如何与更广泛代码库相关的上下文，来协助代码审查。其跨存储库搜索的能力在审查期间为跨项目影响和依赖关系提供了独特的见解。

AI 调试工具

Cody 通过分析错误消息、搜索代码库中的相关代码路径和解释复杂逻辑来帮助开发人员调试问题。其整个存储库的上下文感知能力使其能够跟踪跨多个文件和服务的问题，提供比仅限于单文件上下文的 AI 工具更全面的调试协助。