AI Agent Hub
Back to plugins
🤖

dsh-opencode-zen

Model Inference Updated 2026.08.26

Run the following command in DeepSeek Harness:

dsh plugin install xiaozhe7772222/dsh-opencode-zen

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

Install the dsh-opencode-zen plugin in DeepSeek Harness from its source at https://github.com/xiaozhe7772222/dsh-opencode-zen.

About this plugin

dsh-opencode-zen addresses a simple but common problem: wanting to use free LLMs inside DeepSeek Harness without signing up for accounts, managing API keys, or setting up billing. It connects the OpenCode Zen official public free tier to the DSH model picker, making the opencode provider available with virtually no configuration.

The plugin ships with six free models, including DeepSeek V4 Flash, Xiaomi MiMo, Tencent Hunyuan, NVIDIA Nemotron variants, and Laguna S 2.1. It preserves a full-featured experience with streaming, reasoning-content passthrough, and tool calls, while also adding 429/5xx backoff and request throttling to help stay within free-tier limits. When paired with dsh-api-key-pool, it can automatically rotate across multiple free accounts for more resilient quota usage.

This plugin is a good fit for individual users, lightweight development workflows, and anyone who wants to explore DSH with multiple models at zero cost. If you want a no-signup, no-billing model source for everyday chat, testing, or agent work, dsh-opencode-zen makes that straightforward.

Use Cases

  • Try multiple free LLMs in DSH at zero cost.
  • Use a no-API-key model source for daily chat and light development.
  • Rotate multiple free quotas with an API key pool to reduce rate-limit impact.

Best For

  • DSH users who want a zero-config free model option.
  • Individual developers needing low-cost models for testing.
  • Users who want free models for sessions and sub-agents.