What is Kilo Code?
Kilo Code is an open-source AI coding agent that runs in VS Code, JetBrains IDEs, the terminal, and the cloud. The source code is MIT licensed on GitHub with 26.1K stars and over 25,000 commits.
It works with over 500 models with zero markup on inference costs. You pay what the provider charges. You can bring your own API keys from OpenAI, Anthropic, Google, or other providers, or run models locally through Ollama or LM Studio.
Agent modes
Kilo Code ships with different modes depending on the task:
- Code Mode: writing, refactoring, and shipping production-ready code
- Architect Mode: planning complex features before writing code
- Debug Mode: troubleshooting issues
- Cloud agents: running tasks remotely without your local machine
You can switch between modes depending on what you need.
Pricing
The Kilo Code extension is free and open source. You pay for the AI models you use, with no markup.
Three options for inference:
| Option | Cost | Details |
|---|---|---|
| Auto Free | $0/mo | Routes to free models through Kilo Gateway. Bring your own API key or run local models. |
| Kilo Gateway | $0/mo + usage | Pay-as-you-go through 500+ models across 60+ providers. |
| Kilo Pass | $19/mo | Monthly credit subscription with up to 50% bonus credits. |
You can start without a credit card. No API key required.
For teams, the Teams plan costs $15/user/month with a 14-day free trial. Enterprise pricing is custom.
Where it runs
Kilo Code works across several platforms:
- VS Code extension: install from the Marketplace
- JetBrains plugin: works with all JetBrains IDEs
- CLI: install via npm, curl, brew, or AUR
- Cloud agents: run from your browser
- Slack: team collaboration
- KiloClaw: managed hosting for organizations, starts at $55/month
- Code Reviewer: automated pull request reviews
Key features
- Multi-file code generation from natural language descriptions
- Inline autocomplete with ghost text (tab to accept)
- Self-checking: the agent reviews and corrects its own work
- Terminal and browser control for automating workflows
- BYOK: bring your own API keys from existing providers
- Local model support through Ollama and LM Studio
This post is part of the @aicoding_id series on AI coding tools. Follow us on TikTok for more reviews and tutorials.