dsh-adaptive-model-router
Run the following command in DeepSeek Harness:
dsh plugin install alaxrpg/dsh-adaptive-model-router
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install alaxrpg/dsh-adaptive-model-router inside DeepSeek Harness to install the plugin; the full source is available at https://github.com/alaxrpg/dsh-adaptive-model-router .
About this plugin
In practical DeepSeek Harness workflows, different tasks place very different demands on model capability. A lightweight Q&A prompt and a complex code-reasoning chain need different model strengths and carry different costs. Manually swapping models between tasks is tedious and easy to get wrong, especially when juggling multiple parallel work-streams.
dsh-adaptive-model-router is a model-inference plugin that brings adaptive model routing to Harness, letting the system match the most suitable inference model to each incoming task automatically and trims unnecessary compute spend.
If you routinely handle a mix of inference tasks in DeepSeek Harness and want to cut down on manual model configuration, this plugin is a small but practical addition to your tool-chain.
Use Cases
- Automatically matches each request to the optimal inference model in multi-task parallel workloads
- Eliminates manual model switching when lightweight Q&A and complex code reasoning coexist in the same session
- Dynamically allocates model resources by task complexity in batch inference jobs
Best For
- Developers who run diverse inference tasks in DeepSeek Harness daily
- Teams orchestrating multiple models across parallel work-streams
- Engineering leads focused on inference cost and compute efficiency
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.