DeerFlow

DeerFlow (Kilo) provides an AI coding agent that integrates across various developer environments, including VS Code, JetBrains IDEs (IntelliJ, PyCharm, WebStorm), CLI, a dedicated desktop application (Kilo Desktop for macOS, Windows, Linux), and cloud-based agents.

Users can access over 500 AI models from various providers with zero markup on inference costs. The platform supports bringing your own API keys (BYOK) for providers like Anthropic, OpenAI, Google, Azure, and AWS Bedrock, or running local models with Ollama or LM Studio for enhanced privacy.

DeerFlow (Kilo) includes specialized agent modes to address different stages of the development workflow. These include Code Mode for writing and refactoring, Architect Mode for planning complex features, and Debug Mode for identifying and fixing bugs.

Users can run AI agents in the cloud, allowing for the execution of long-running tasks, complex workflows, and resource-intensive operations without consuming local machine resources. This enables developers to keep their local environment free for other work.

The platform offers automated code review agents that analyze pull requests, provide suggestions for improvements, identify potential bugs, and help ensure adherence to code quality standards, offering instant feedback before code merges.

DeerFlow (Kilo) allows users to start an AI agent session on one platform or device and seamlessly continue it on another. This ensures that AI agents, context, and progress follow the user across different environments, maintaining workflow continuity.

The platform is open-source under an MIT license, providing full transparency. Users can inspect the prompt, context window, and decisions made by the AI. It is also forkable, modifiable, and self-hostable, allowing for customization and community contributions.