Free Kimi-k3 in JetBrains: Configure Continue in 5 Minutes

The article from Habr describes how to set up the Kimi-k3-free model in JetBrains IDEs using the Continue plugin. Kimi K3, developed by Moonshot AI, has 2.8 trillion parameters, a 1 million token context, and native image support. The setup involves registering on TokenRouter, creating an API key, installing the Continue plugin from the JetBrains plugin repository, and configuring it with a JSON snippet that specifies the model, provider, API base, and roles. The author notes that free access is currently available and includes limitations; they also share their personal configuration which adds local models (qwen3-coder, gemma4) via Ollama, DeepSeek V4 flash and pro models via the DeepSeek API, as well as custom prompts and rules. The article advises restarting the IDE after changes to Continue to avoid lag and suggests invalidating the cache.