AI Agent Hub
Back to plugins
🤖

dsh-model-router

Model Inference Updated 2026.08.24

Run the following command in DeepSeek Harness:

dsh plugin install superboy911/dsh-model-router

Paste the following prompt into your AI chat to install this plugin:

To install the dsh-model-router plugin in DeepSeek Harness, run the command dsh plugin install superboy911/dsh-model-router, with source code available at https://github.com/superboy911/dsh-model-router.

About this plugin

When using DeepSeek Harness for model inference, you might want conversations to automatically route to the most appropriate model without always relying on manual selection. The dsh-model-router plugin is designed for this: it allows you to set up deterministic keyword routing, so mentioning 'design' can automatically switch to Kimi K3, or upgrade models based on task complexity. Core capabilities include a default-off allowlisted model_route tool for safe switching to specified models, and an isolated image generation channel for secure calls to models like Doubao Seedream 5.0 Lite. This plugin is ideal for all DeepSeek Harness users, from daily Q&A to specialized tasks, boosting efficiency and streamlining workflows; it also invites developer contributions to refine routing strategies.

Use Cases

  • Automatically select the best model based on keywords in conversation
  • Safely invoke the isolated image generation channel to create visual content
  • Optimize model usage costs via the allowlisted tool

Best For

  • Users of DeepSeek Harness for model inference
  • Developers needing custom routing strategies to boost efficiency
  • Teams focused on cost control and task automation