dsh-auto-model-router
Run the following command in DeepSeek Harness:
dsh plugin install ADVeRTs13/dsh-auto-model-router
Paste the following prompt into your AI chat to install this plugin:
Install this plugin in DeepSeek Harness using the command: dsh plugin install ADVeRTs13/dsh-auto-model-router. Source: https://github.com/ADVeRTs13/dsh-auto-model-router
About this plugin
Balancing inference cost and response quality is a common challenge when using DeepSeek Harness. The dsh-auto-model-router plugin introduces a hybrid auto-routing mechanism that classifies tasks into four capability tiers (L0-L3). It intelligently routes requests to the most suitable model based on complexity, utilizing keyword scoring or optional LLM classification. With features like zero-latency heuristics and built-in budget controls, it ensures optimal performance for everything from casual conversation to complex code refactoring, helping you manage expenses without sacrificing results.
Use Cases
- Automatically allocates models for both daily chat and complex coding tasks.
- Wants to reduce API costs while maintaining response quality.
- Needs a unified strategy to manage model routing across different capability levels.
Best For
- DeepSeek Harness users concerned about inference costs.
- Developers requiring automated model management.
- Advanced users seeking flexible model switching based on task complexity.
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.