Cody
Cody is Sourcegraph's AI coding assistant. Official docs say it is supported on Sourcegraph Enterprise and available in VS Code, JetBrains, Visual Studio, the Sourcegraph web app, and a CLI. It uses current LLMs plus Sourcegraph Search so chat and edits can pull APIs, symbols, and usage patterns from local and remote repositories on GitHub or GitLab. That codebase graph is the product, not a generic autocomplete plugin.
Key Features
- Chat with
@context: Open file and repo are in context by default.@adds files, symbols, remote repos, or non-code artifacts. - Auto-edit: After one character edit, Cody proposes contextual changes from cursor movement and recent typing.
- Prompts: Premade and custom prompts for repeated team tasks.
- Enterprise search backend: Docs emphasize Sourcegraph Search across the whole codebase, not only the open tab.
- MCP, API, and CLI on the platform: Sourcegraph pricing also lists MCP, GraphQL/REST, and CLI so agents such as Claude Code, Cursor, Codex, and Amp can reuse the same code graph.
Use Cases
- Monorepos and multi-repo companies: Ask about an API that lives in another service.
- Enterprise rollouts: VS Code, JetBrains, Visual Studio, and the web app under one Sourcegraph deployment.
- Agent context: Point other coding agents at Sourcegraph MCP instead of stuffing the repo into a prompt.
Pricing
Sourcegraph's public pricing page (accessed 2026-08-17) no longer lists a self-serve Cody Free/Pro grid. The commercial offer is an Enterprise plan that "starts at $16K", includes credits for AI features, and scales with team size. Contact sales for seats, self-hosted versus single-tenant cloud, and 24x5 support. Treat older "free unlimited autocomplete" claims as outdated.
Getting Started
- Confirm you have a Sourcegraph Enterprise workspace, or talk to sales.
- Install Cody for VS Code or JetBrains. Visual Studio is marked experimental.
- Connect GitHub or GitLab so remote context resolves.
- Start with chat on the current file, then add
@remote repos before you trust auto-edit.
Frequently Asked Questions
Is Cody still a consumer plugin?
Docs now frame it as an Enterprise product with IDE and web clients. Do not plan a free individual tier without a current Sourcegraph quote.
Does it work without Sourcegraph?
Official copy ties Cody to Sourcegraph Search and Enterprise. The interesting part is that graph, not the chat UI alone.
How is it different from Copilot?
GitHub Copilot lives in GitHub. Cody is built to search and edit across the Sourcegraph index.
Alternatives
- GitHub Copilot: Default if the company already standardizes on GitHub.
- Cursor: Editor-native agent for smaller teams.
- Continue: Open-source IDE agent if you cannot buy Sourcegraph.
Tips
- Use
@remote context early. Local-file chat wastes the Search API. - Turn custom prompts into team defaults instead of pasting the same spec.
- Recheck the $16K Enterprise starting point on sourcegraph.com/pricing before you write a budget.
Conclusion
Cody is for teams that already need Sourcegraph-scale code search and want an assistant on top of that index. It is not the 2023 free autocomplete sidebar. Start from the Cody docs and a current Enterprise quote.
Comments
No comments yet. Be the first to comment!
Related Tools
GitHub Copilot
github.com/features/copilot
GitHub's AI coding assistant for IDE, CLI, mobile, and cloud agents. Completions, chat, PR drafts, and paid plans from Free to Enterprise.
aiXcoder
www.aixcoder.com
Chinese AI coding assistant with an Apache-2.0 7B model. No public SaaS price table; treat cloud seats as sales-quoted.
Amp
ampcode.com
Sourcegraph agent. Rechecked: Megawatt $20 / Gigawatt $200. Not a Cody rename; not $10/day unconstrained Opus.
Related Insights
Skills + Hooks + Plugins: How Anthropic Redefined AI Coding Tool Extensibility
An in-depth analysis of Claude Code's trinity architecture of Skills, Hooks, and Plugins. Explore why this design is more advanced than GitHub Copilot and Cursor, and how it redefines AI coding tool extensibility through open standards.
Seven AI Coding CLIs, Six Months: No Matter How Strong the Model, Work Needs Supervision
Claude Code, Codex, opencode, pi, omp and DeepSeek Harness all have personalities. After six months of deep use I run a division of labor: pi for the fastest cheapest reviews, omp for complex PRs, DeepSeek Harness on V4 Flash for high-frequency low-cost review, and Claude Code, Qoder and Cursor for writing. No matter how strong the model, work needs supervision — ideally from an independent third party.
Claudesidian: Transform Obsidian into an AI-Powered Second Brain
Discover Claudesidian, an open-source project that perfectly integrates Obsidian with Claude Code. Built-in PARA method, custom commands, and automated workflows for a complete idea-to-implementation solution.