AI models recommended for developers by the opencode team
opencode is an open-source AI coding agent built for the terminal. It features a native TUI, multi-session support, and is model-agnostic, letting you use Claude, OpenAI, Google, or even local models.
OpenCode Zen simplifies onboarding by offering a curated list of pre-tested models; users can run the /connect command, sign in at opencode.ai/auth, add billing details, and paste their API key to immediately access validated models optimized for coding tasks.
OpenCode Zen provides a handpicked set of AI models that have been tested and benchmarked by the OpenCode team specifically for coding agents, eliminating the need to evaluate and validate models independently while ensuring consistent quality and performance.
OpenCode Zen integrates with OpenCode, which is available as a terminal interface, desktop app (beta on macOS, Windows, Linux), and IDE extensions (VS Code, Cursor, Zed, Windsurf), allowing developers to access the same validated models across their preferred development environment.
OpenCode Zen operates through OpenCode, which does not store any code or context data, allowing all processing to happen locally or through direct API calls to the user's chosen AI provider, making it suitable for privacy-sensitive environments and organizations with strict data governance requirements.
OpenCode Zen allows users to log in with GitHub Copilot, OpenAI ChatGPT Plus/Pro, or connect to any model from 75+ LLM providers through Models.dev, including local models, providing flexibility to use existing subscriptions or preferred providers while benefiting from Zen's pre-validated model selection.