Helm API

Helm API

English · Chinese

One control plane for LLM traffic: text, images, subscriptions, fallback, and memory.

Open-source · self-hosted · MIT

LLM apps tend to accumulate routing code in all the wrong places: fallback lists inside clients, one-off patches for provider quirks, hard-coded model names, ad hoc cost controls, and no clean way to answer "why did this request go there?"

Helm API puts that work in one place: an open-source, self-hosted LLM routing gateway — *nginx for the LLM world*. Your app sends normal OpenAI, Anthropic, Gemini, or image-generation requests. Helm classifies the request, picks a lane, chooses a provider account, falls back when an upstream breaks, translates protocols when needed, and records the whole decision trail. Clients usually change only base_url and the API key.
Manage traffic as configuration, not as code.

Change the model behind a lane? Edit one YAML line — or click in the dashboard. Your apps never notice.

The dashboard — live traffic, token usage by model, spend, and the most recent routing decisions.
Screenshot note: the Admin images were captured on 2026-07-05 from
v0.25.2 and are retained as historical layout examples. The prose in this
README reflects the current source; screenshot labels, data windows, and the
visible version badge may differ. See 11 · Admin UI for
full provenance.

Quickstart

Prerequisites: Docker, or Node ≥ 22 + pnpm 10 to build from source.

Open the complete /setup#token=... URL printed by the script. The protected
token stays in the browser URL fragment and is consumed automatically—there is
nothing to paste. The wizard starts with the Admin login, then lets you test and
save optional provider keys, or continue with no API key and connect a
ChatGPT/Codex, Claude, Copilot, or Grok subscription…