Cody
Sourcegraph's AI assistant powered by code graph indexing for multi-repo enterprise context.
Overview & Analysis
Cody is Sourcegraph's AI coding assistant. Powered by Sourcegraph's code graph indexing technology, Cody excels at answering complex questions and writing code across large, multi-repository codebases where standard single-file assistants lose context.
Pricing Model: Free tier with 200 completions/month; Pro at $9/month; Enterprise at $19/user/mo.
Free Tier Details: 200 completions and 20 chat messages per month on the free tier.
Technical Specs & Models
Key Capabilities & Features
Specialized engineering mechanisms implemented in Cody.
Multi-Repository Code Graph
Searches and indexes across your entire organization's GitHub and GitLab repositories.
Model Flexibility
Choose between Anthropic Claude 3.5 Sonnet, OpenAI GPT-4o, and Google Gemini.
Affordable Pro Tier
Pro tier at only $9/month offers unlimited chat and completions with frontier models.
Pros & Cons Evaluation
Forensic analysis of developer experience, cost efficiency, and trade-offs.
- Exceptional multi-repository search powered by Sourcegraph's proven search engine
- Pro plan at $9/month is the most affordable paid assistant on the market
- Supports VS Code, JetBrains IDEs, and Neovim
- Multi-file agentic editing is slower than Cursor Composer or Windsurf Cascade
Pricing Plans & Tiers
Subscription plans, credit quotas, and commercial deployment options.
Free
For individual developers
- 200 completions/month
- 20 chat messages/month
- Standard multi-repo context
Pro
For professional software engineers
- Unlimited completions & chats
- Choice of Claude 3.5 Sonnet & GPT-4o
- Advanced codebase indexing
- Priority support
Enterprise
For enterprise engineering teams
- Full Sourcegraph code graph integration
- Custom model endpoints & BYOK
- Single Sign-On (SSO)
- Audit logging & zero retention
Verified Benchmark Evaluations
Standardized evaluation scores recorded on SWE-bench and synthetic code harnesses.
| Benchmark Harness | Score / Accuracy | Evaluation Metric | Verification Status | Evaluation Notes |
|---|---|---|---|---|
| DeepSWE v1.1 | 63.5% | Task Completion | Verified | Standard benchmark instance run. |
| SWE-Atlas | 72.5% | Multi-Repo Search | Verified | Standard benchmark instance run. |
Alternative AI Coding Agents & Comparisons
Compare Cody against other top-ranked software engineering assistants.
The world's most widely adopted AI developer tool, deeply embedded across Microsoft's development ecosystem.
Enterprise AI coding platform built for massive million-line codebases and complex multi-repo architectures.
The leading open-source AI code assistant for VS Code and JetBrains with full local model support.