What Is Kilo Code?
Kilo Code is an open source AI coding platform designed around model freedom and consistent agent workflows across development surfaces. The same core agent can be used in supported IDEs from the command line and through cloud based development tasks.
Hundreds Of Models Without One Vendor Lock In
Kilo provides access to a large model catalog through its Gateway and also allows developers to bring their own provider credentials or configure local models. Its pricing model emphasizes direct provider rates rather than adding a hidden inference markup and a free routing option can make basic use possible without paid model spend.
Modes For Different Engineering Tasks
Specialized agent modes separate implementation architecture and debugging work. Teams can move between local and cloud agents while code review and collaboration features extend the workflow beyond a single developer session.
Best For
- Developers who want broad AI model choice
- Teams using both VS Code and JetBrains environments
- Users who prefer open source coding infrastructure
- Cost conscious developers using free BYOK or local models
- Engineering teams experimenting with cloud coding agents
Key features
- 500 Plus Models: Connects a very large catalog of model providers.
- Bring Your Own Key: Uses external provider credentials without locking users into one service.
- Local Model Support: Enables private or offline compatible model workflows.
- VS Code And JetBrains: Supports major professional IDE families.
- Kilo CLI: Runs the coding agent from a terminal.
- Cloud Agents: Executes development work remotely.
- Specialized Modes: Includes coding architecture and debugging workflows.
- Kilo Gateway: Provides unified access and routing across many models.
- Open Source: Allows code inspection modification and self managed workflows.
Use cases
AI Coding,Model Comparison,Feature Development,Debugging,Architecture Planning,Cloud Agents,Local Model Coding,Team Code Review
How to use
How To Use Kilo Code
- Choose A Surface: Install Kilo in VS Code or JetBrains use the CLI or start from a supported cloud workflow.
- Select Model Access: Use Kilo Auto choose a Gateway model add your own API key or configure a compatible local model.
- Open The Repository: Give the agent access to the project that needs work.
- Pick An Agent Mode: Use Code for implementation Architect for planning or Debug for investigation.
- Describe The Task: Provide the expected result constraints and any files or tests that matter.
- Review Agent Actions: Check code edits commands and model usage as the session progresses.
- Test The Result: Run the project's normal checks and inspect the generated diff.
- Move Work To The Right Surface: Use cloud agents review tools or team features when the task needs asynchronous collaboration.
Best for
Developers, Teams, Model Flexible Users, Open Source Users, Cost Conscious Coders
Integrations
VS Code,JetBrains,CLI,Cloud Agents,Slack,MCP,Kilo Gateway,Local Models
Commercial use
Depends On Kilo Terms Open Source License And Selected Model Provider Terms
