dsh-llm-subscription
Run the following command in DeepSeek Harness:
dsh plugin install DavidRm1911/dsh-llm-subscription
Paste the following prompt into your AI chat to install this plugin:
Run the install command in DeepSeek Harness to enable the plugin; the source is at https://github.com/DavidRm1911/dsh-llm-subscription .
About this plugin
The DeepSeek Harness model picker shows only DeepSeek models by default. If you already pay for a Claude Code or Antigravity subscription and would rather not provision a separate metered API key, dsh-llm-subscription fills that gap: it registers through the DSH native dsh-llm adapter seam, placing Claude, Gemini, and a local Ollama model directly into the model picker. No Custom Provider screen, no sidecar process to keep alive — it is simply there once installed.
Three provider paths are covered. The Claude side exposes claude, haiku, sonnet, opus, and fable, each with a native reasoning-effort selector ranging from Low to Max. The Gemini side registers five Antigravity tiers (Flash Low/Medium/High, Pro Low/High). A local Ollama Qwen model is also registered as a zero-cost, zero-login fallback. Every response ends with a via line that names the model which actually produced the visible text, so there is no guesswork about which model answered.
Best suited for developers who already maintain a Claude Code or Antigravity subscription and want to switch models inside a DSH workflow without extra API spend, as well as users who prefer a local Ollama model for cost-free inference. Keep in mind: this is an unofficial, community-built plugin; routing requests through an existing subscription carries a real account-ban risk under Anthropic and Google terms of service. Streaming deltas and chain-of-thought text are not currently available. Review the vendor usage policies and assess the risk yourself before installing.
Use Cases
- Pick Claude Sonnet or Opus directly from the DSH model picker for coding tasks using an existing subscription
- Switch between Gemini Flash and Pro tiers via Antigravity for multi-turn dialogue and reasoning
- Invoke a local Ollama Qwen model for zero-login, zero-cost offline inference
Best For
- Developers with an existing Claude Code subscription who want to call Claude inside DSH without an extra API key
- Users on an Antigravity subscription who need seamless Gemini access within a DSH workflow
- Developers who prefer local Ollama models for zero-cost, privacy-preserving inference
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.