Hacker News Reader

Does GLM-5 have multimodality or are they still wanting you to load an MCP for vision support?

Text only still, sadly, though qwen3.5-plus on the same provider (Model Studio) is