wenzetan/dsh-llm-newapi
NewAPI (OpenAI-compatible gateway) LLM provider: registers a newapi route with chat-only model discovery, auto-fills model parameters (context window, reasoning effort) from models.dev, and adds a Web settings section for the base URL and API key.
dsh-llm-newapi adds NewAPI as a DeepSeek Harness LLM provider with zero modifications to dsh itself: provider route id 'newapi', implementing the LlmAdapter seam from @deepseek-ai/dsh-llm. It is dual-face: a host side (adapter + model discovery) plus a browser side — a 'NewAPI' settings page in the dsh web settings panel with 'Fetch model info'. After install, fill in the API key and gateway address (including /v1), fetch the model list, and the newapi route's chat models appear in the model picker (embedding/rerank/ranker models are filtered automatically).
Install
dsh plugin --profile web add dsh-llm-newapinpm dsh-llm-newapi 0.8.4 verified 2026-09-04 (repo field empty on npm but maintainer wenzetan == GitHub repo owner wenzetan/dsh-llm-newapi — owner-consistent; README EN primary w/ zh edition). Dual release channel: dsh plugin --profile web add dsh-llm-newapi (npm latest = last manually confirmed stable) or dsh plugin --profile web add dsh-llm-newapi@next (prerelease); never install main HEAD. Requires dsh >=0.1.0-rc.
Compatibility
DeepSeek Harness >=0.1.0-rc; NewAPI is an OpenAI-compatible gateway (POST {baseURL}/chat/completions, GET {baseURL}/models, baseURL includes /v1).
Details
- Repo: wenzetan/dsh-llm-newapi
- Category: Models & Providers
- Stars: 0
- Version: npm dsh-llm-newapi 0.8.4
- Last push: 2026-09-16
- First seen: 2026-08-14
Recent updates
Release channels: stable (npm latest / github #latest), prerelease (@next / #vX.Y.Z-rc.N), main HEAD not recommended.
FAQ
- What is NewAPI?
- An OpenAI-compatible API gateway: POST {baseURL}/chat/completions and GET {baseURL}/models, with the baseURL including /v1. This plugin registers it as a 'newapi' provider route in dsh.
- Which channel should I install?
- The stable channel — dsh plugin --profile web add dsh-llm-newapi resolves npm latest, which always points at the newest manually confirmed stable version. Use @next only to test prereleases; never install main HEAD.
- How do I add models after installing?
- Open Settings → NewAPI, enter the API key and gateway address (including /v1), click 'Fetch model info', then pick chat models — embedding/rerank/ranker models are filtered automatically.
Alternatives
wenzetan/dsh-quota-panel · wss534857356/dsh-plugin-codex