Skip to main content

LLM Configuration

LLM configurations link a model with a credential — telling the platform which AI model to use and how to authenticate with its provider.

Models​

ORQO supports models from multiple providers. Models are synced from provider APIs, so the available list stays current as providers release new models.

Supported providers include:

  • OpenRouter — Access to 200+ models through a single API key (recommended).
  • OpenAI — GPT-4.1, GPT-4o, o1, o3, and more.
  • Anthropic — Claude Opus, Sonnet, Haiku.
  • Google — Gemini models.
  • Ollama — Run open-source models locally or via Ollama Cloud. No API key needed for local inference.
  • LM Studio — Local models via OpenAI-compatible API.
  • Custom — Any OpenAI-compatible endpoint (vLLM, TGI, corporate gateways).

Full details on each provider, setup instructions, and model recommendations: LLM Providers

On ORQO's included model access, models are routed only to providers operating under a Zero Data Retention policy. Models that qualify show a ZDR badge in the picker, and a ZDR only filter narrows the list to just those models. See Data Privacy.

When you choose a model, the model picker shows its pricing tier, capabilities, context window, and data-handling at a glance — see The Model Picker for what every column, icon, and badge means.

System Fallback — No API Key Required​

Every LLM configuration has a built-in system fallback. If you don't assign your own credential, ORQO automatically uses its platform key for that provider. Usage is deducted from your subscription's credit balance.

Platform keys are available per provider — OpenAI, Anthropic, Google, and OpenRouter each have their own. This means you can start building and running workflows immediately — no API keys needed.

The platform resolves credentials through a simple chain:

  1. Your credential — If you've assigned one to the LLM config, it's used directly.
  2. Platform key — If no credential is assigned, the platform's own key for that provider handles the request and bills your subscription credits.

Creating a Custom LLM Configuration​

To use your own API key for lower costs or higher rate limits, go to Settings → LLM Configs:

  1. Select a model from the synced list.
  2. Select the credential that holds your API key for that provider (e.g., your own OpenRouter, OpenAI, or Anthropic key).
  3. Give it a descriptive name (e.g., "Claude Sonnet - Production").
tip

Using your own OpenRouter key is the most flexible option — a single key gives you access to all 200+ models without switching providers.

Assigning LLM Configs​

LLM configurations are assigned at two levels:

  • Team level — All agents in the team use this config by default.
  • Agent level — An individual agent can override the team's config with a different one.

This lets you mix models within a team. For example, use a fast model for research agents and a more capable model for the final review agent.