Last released May 15, 2026
Vision-empowered MCP server for OpenCode text-only models. PaddleOCR (SOTA deep learning) + Google Gemini fallback for model-agnostic image analysis.
Supported by
