dsh-pro-vision
Run the following command in DeepSeek Harness:
dsh plugin install lasdrder0705/dsh-pro-vision
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install lasdrder0705/dsh-pro-vision in your terminal to install; the source code lives at https://github.com/lasdrder0705/dsh-pro-vision. Restart dsh web after installation to activate the plugin.
About this plugin
DeepSeek-V4-Pro is a text-only reasoning model. The moment a user drops a screenshot, a chart, or a product photo into the conversation, Pro has no way to interpret it. Yet in day-to-day workflows a single error screenshot often conveys more than ten lines of typed description.
dsh-pro-vision fills that gap without ever swapping your primary model. Before the message reaches Pro, the plugin intercepts attached images, sends them to deepseek-v4-flash-vision-exp for visual understanding, and splices the resulting caption back into the prompt as an [Image N] block. Each attachment is cached in-process so repeated turns do not re-invoke the vision model; and if you explicitly select a vision model, the plugin stays completely out of the way.
Built for developers who already rely on V4-Pro for long-context reasoning and simply want to paste an image now and then without switching profiles. Pure JavaScript, zero native dependencies, one install command across macOS, Windows, and Linux.
Use Cases
- Paste screenshots or charts into a text-only reasoning session and let Pro interpret them inline
- Handle tickets containing product photos or error captures without swapping to a vision model
- Reference the same attached image across multiple turns with in-process caching to avoid redundant vision calls
Best For
- Developers who rely on DeepSeek-V4-Pro for long-context reasoning and occasionally need image input
- Users who want to mix text and images in a single workflow without managing separate model profiles
- Teams that prefer pure-JS, zero-native-dependency, cross-platform plugins with a one-command install
Related Plugins
ModLens is a vision plugin for DeepSeek Harness that gives text-only models sight by reading images pasted directly into chat, with zero-config setup and multiple vision engines.
On-demand vision for text-only DeepSeek Harness agents: built-in free keyless vision chain and 14 vision tools, routing image turns as tool calls to vision models with pixel fidelity, no Python needed, one-command install.
Give text-only models in DeepSeek Harness eyes, enabling image Q&A, long-screenshot OCR, UI restoration, and GUI visual tasks.
An armor-breaking plugin for DeepSeek: rewrites instruction priority, expands refusal-blocked phrases, persists across sessions, and shows a green active indicator.