Skip to content

providers.md

Latest commit

 

History

History
230 lines (175 loc) · 4.17 KB

providers.md

File metadata and controls

230 lines (175 loc) · 4.17 KB

Provider Setup Guide

Quorum supports any provider available through the pi-ai library. This guide covers setup for each supported provider.

Auto-Detection

The fastest way to get started:

export OPENAI_API_KEY=sk-...
export ANTHROPIC_API_KEY=sk-ant-...
quorum init

Quorum scans environment variables and local services, then configures providers automatically.

Cloud Providers

OpenAI

export OPENAI_API_KEY=sk-...

Get your key at platform.openai.com/api-keys.

Config:

- name: openai
  provider: openai
  model: gpt-4o  # or gpt-4o-mini, o3-pro, etc.
  auth:
    method: env
    envVar: OPENAI_API_KEY

Anthropic (Claude)

export ANTHROPIC_API_KEY=sk-ant-...

Get your key at console.anthropic.com.

Config:

- name: claude
  provider: anthropic
  model: claude-sonnet-4-20250514
  auth:
    method: env
    envVar: ANTHROPIC_API_KEY

OAuth (Claude Code): If you use Claude Code with OAuth, Quorum can read the token from macOS Keychain:

- name: claude
  provider: anthropic
  model: claude-sonnet-4-20250514
  auth:
    method: oauth_keychain
    service: com.anthropic.claude-code

Google (Gemini)

export GOOGLE_GENERATIVE_AI_API_KEY=AI...

Get your key at aistudio.google.com/apikey.

Config:

- name: gemini
  provider: google
  model: gemini-2.0-flash
  auth:
    method: env
    envVar: GOOGLE_GENERATIVE_AI_API_KEY

Gemini CLI: If you have Gemini CLI installed, Quorum can detect it automatically.

Kimi (Moonshot)

export KIMI_API_KEY=sk-...

Get your key at platform.moonshot.cn.

Config:

- name: kimi
  provider: kimi
  model: moonshot-v1-auto
  auth:
    method: env
    envVar: KIMI_API_KEY

DeepSeek

export DEEPSEEK_API_KEY=sk-...

Get your key at platform.deepseek.com.

Config:

- name: deepseek
  provider: deepseek
  model: deepseek-chat
  auth:
    method: env
    envVar: DEEPSEEK_API_KEY

Mistral

export MISTRAL_API_KEY=...

Get your key at console.mistral.ai.

Config:

- name: mistral
  provider: mistral
  model: mistral-large-latest
  auth:
    method: env
    envVar: MISTRAL_API_KEY

Groq

export GROQ_API_KEY=gsk_...

Get your key at console.groq.com.

Config:

- name: groq
  provider: openai
  model: llama-3.3-70b-versatile
  baseUrl: https://api.groq.com/openai/v1
  auth:
    method: env
    envVar: GROQ_API_KEY

Note: Groq uses OpenAI-compatible API, so set provider: openai with a custom baseUrl.

Local Providers

Ollama

Install from ollama.com, then:

ollama pull llama3

No API key needed. Quorum auto-detects Ollama at http://localhost:11434.

Config:

- name: ollama
  provider: ollama
  model: llama3
  auth:
    method: none

LM Studio

Run LM Studio's local server (default: http://localhost:1234). Quorum auto-detects it.

Config:

- name: lmstudio
  provider: ollama
  model: your-model-name
  baseUrl: http://localhost:1234
  auth:
    method: none

Custom / OpenAI-Compatible

Any OpenAI-compatible API:

- name: my-provider
  provider: openai
  model: my-model
  baseUrl: https://my-api.example.com/v1
  auth:
    method: env
    envVar: MY_API_KEY

Managing Providers

quorum providers list           # show all configured
quorum providers test           # test connectivity
quorum providers models         # browse available models
quorum providers add --name X --type openai --model Y --env Z
quorum providers remove X

Recommended Minimum Setup

For best deliberation quality, configure at least 3 providers from different families:

export ANTHROPIC_API_KEY=sk-ant-...
export OPENAI_API_KEY=sk-...
export GOOGLE_GENERATIVE_AI_API_KEY=AI...
quorum init

This gives you Claude + GPT + Gemini — three distinct model architectures that produce genuinely diverse perspectives.