AI Agent Hub
Back to plugins
🤖

dsh-model-sync

Model Inference Updated 2026.08.27

Run the following command in DeepSeek Harness:

dsh plugin install GooDAnDReaDY/dsh-model-sync

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

Run dsh plugin install GooDAnDReaDY/dsh-model-sync in your DeepSeek Harness terminal to install the plugin; the full source code is hosted at https://github.com/GooDAnDReaDY/dsh-model-sync.

About this plugin

Every time an upstream AI provider ships a new model, tweaks a context window, or adjusts pricing, hand-editing YAML configs is slow and error-prone. dsh-model-sync automates the entire workflow: it polls 25+ major providers (OpenAI, Anthropic, Google, DeepSeek, xAI, OpenRouter, Groq, Mistral, Ollama, and more) to discover newly released and retired models, synchronizes capability flags such as vision, tools, reasoning, and embeddings, and queries upstream billing endpoints to track account credits so your service never blindsides into an exhausted quota. Every change is recorded as a timestamped diff in the Web dashboard, and you can trigger an on-demand refresh, toggle providers on or off, or plug in any custom OpenAI-compatible gateway.

The plugin stays reliable at scale. Whether you manage a handful of endpoints or a catalog with ten thousand or more models, chunked rendering, iterative aggregation, deterministic diff serialization, adaptive Retry-After backoff, and automatic SETTINGS_CONFLICT recovery keep both the background sync and the UI responsive and crash-free. It is a strong fit for teams and individual developers running DeepSeek Harness across multiple AI providers who want a hands-off catalog synchronization layer, real-time balance monitoring, and a fully auditable sync history without writing custom scripts.

Use Cases

  • Auto-sync model catalogs and capability flags across multiple AI providers on every upstream release
  • Track upstream account credits and quotas to prevent unexpected service outages
  • Audit model deprecations, additions, and capability changes with timestamped diffs

Best For

  • DeepSeek Harness operators integrating multiple upstream AI providers
  • Mid-to-large LLM platform teams needing hands-off catalog synchronization
  • Fin-ops teams responsible for quota tracking, cost control, and compliance auditing