modlens
Docs文档

Getting started

ModLens is a CLI plugin that reads images for text-only models. An image goes in. Structured evidence comes out.

The flagship DeepSeek and GLM chat models cannot read images. ModLens is a CLI they can call. It reads images pasted straight into the chat. You do not have to save a file and pass a path first.

It is for people running text-only DeepSeek, GLM, or MiMo Pro inside Claude Code, Codex, Pi, OpenCode, or DeepSeek Harness.

Config lives in ~/.modlens/config.json. The website is not in the runtime.

Install on DeepSeek Harness

On dsh, ModLens is a native plugin:

npx -y @deepseek-ai/dsh plugin --profile web add @liustack/modlens@3.24.2

That registers the modlens_read_image tool. Restart dsh and look for (modlens vision) in the model selector.

Install on other harnesses

Send this line to your agent:

Install and configure the modlens skill following https://github.com/liustack/modlens/blob/main/INSTALL.md, then run the health check and tell me the result.

The procedure is on the Install page. It checks what the machine already has, asks before reusing a local login, and finishes with modlens doctor.

First read

Once a vision engine is ready:

modlens -i screenshot.png

Output is a fixed JSON shape. Flags and the failover chain are in the CLI manual. The JSON contract is in Output schema.

Next