Skip to main content
Continuum runs OpenRouter models through a shared opencode serve process. This is the bring-your-own-API-key path.

Requirements

opencode must be installed. Authenticate your OpenRouter (or other supported) API key with:
opencode auth login
Continuum reads the OpenCode auth file and starts opencode serve automatically.

How it integrates

Continuum:
  • Starts a shared opencode serve process on the Mac.
  • Sends prompts through OpenCode’s localhost HTTP API.
  • Consumes the SSE event stream that opencode serve emits.
  • Maps OpenCode usage events into the Continuum analytics layer.
  • Surfaces live model metadata under Settings → Providers.
OpenCode sessions appear in the Chat tab, the Code tab session list, and the Usage tab.

Usage analytics

OpenRouter does not expose Anthropic-style rolling quota headers. Continuum tracks OpenRouter usage as a dollar-cost lane in the Usage tab rather than a percentage-based quota gauge. Costs are calculated from token counts returned by the SSE stream and the LiteLLM pricing snapshot.

Model metadata

Available OpenRouter models are fetched from opencode serve and shown in Settings → Providers. The model picker in the Chat and Code tabs is populated from this live list.

Broadcast

OpenRouter via OpenCode can participate in the Chat tab’s multi-provider broadcast mode alongside Claude, Codex, Antigravity, Cursor, and Grok.
Provider CLI telemetry is owned by OpenCode and its upstream providers, not Continuum. Continuum consumes OpenCode’s localhost SSE stream and has no visibility into OpenCode’s upstream API calls to OpenRouter or other providers. See Privacy for the full egress breakdown.