dsh-model-search
Run the following command in DeepSeek Harness:
dsh plugin install fanfan6/dsh-model-search
Paste the following prompt into your AI chat to install this plugin:
Install this plugin by running dsh plugin install fanfan6/dsh-model-search in DeepSeek Harness (full source at https://github.com/fanfan6/dsh-model-search), then restart DSH to see the search button in the input toolbar.
About this plugin
When you run several model providers in DeepSeek Harness - SenseTime, Bailian, Xiaomi, local endpoints - switching models means scrolling through a long, mixed list and spotting the right entry by eye. dsh-model-search replaces that with a search-first workflow: a toolbar button opens a panel, you type a keyword, every configured source is filtered in real time, and one click applies the model to the current session.
The filter matches on both platform name and model name. Type qwen and you see every entry containing that string across all providers; type Bailian and the list narrows to a single platform. Results render as compact single-line rows (platform model-name) so you can scan and confirm a pick in seconds. The panel inherits DSH CSS variables and adapts automatically to light or dark themes with zero extra configuration.
If you swap between three or more providers on a daily basis, or any single provider exposes a dozen-plus models, this plugin turns model selection from a scrolling exercise into a quick search-and-click gesture, keeping your focus on the conversation rather than the settings menu.
Use Cases
- Managing multiple providers (SenseTime, Bailian, Xiaomi) and needing to locate a specific model instantly
- A single provider exposes a dozen-plus models and scrolling through the full list is tedious
- Frequent cross-platform model switching where a single search-and-click replaces repeated menu navigation
Best For
- Developers running three or more model providers in DSH
- Researchers and engineers who switch between platforms multiple times daily
- DSH users who want to minimize friction in model selection and keep their conversation flow unbroken
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.