AI Agent Hub
Back to plugins
🤖

dsh-polyglot

Model Inference Updated 2026.08.21

Run the following command in DeepSeek Harness:

dsh plugin install Jesse-njx/dsh-polyglot

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

To install this plugin in DeepSeek Harness, run the command dsh plugin install Jesse-njx/dsh-polyglot or visit the full source address at https://github.com/Jesse-njx/dsh-polyglot for more details.

About this plugin

Dsh-polyglot acts as a model switch for DSH, addressing the common issues of rate limits and failures when working with multiple OpenAI-compatible endpoints. It automatically fails over to alternative providers when free tiers are rate-limited or when a provider goes down, ensuring uninterrupted service.

Its core capabilities include a generic adapter that handles streaming, tool calls, and usage extraction, along with a router with exponential backoff for seamless fallback. Presets are managed as data, allowing the community to add providers easily, and usage visibility helps users monitor costs and performance.

This plugin is ideal for DSH users such as developers, researchers, or anyone relying on model inference. If you frequently use free tiers or need a reliable backup, dsh-polyglot lets you focus on your work without worrying about backend stability.

Use Cases

  • Automatically switch to alternative providers when free model services are rate-limited.
  • Easily manage multiple OpenAI-compatible endpoints through presets.
  • Monitor model usage and costs in real-time.

Best For

  • Developers and researchers using DSH for model inference.
  • Users who need reliable failover and backup options.
  • Individuals relying on free or cheap model providers and focusing on cost control.