Observed facts: Vercel announced Beam availability through AI Gateway with free use for a limited time during launch when serviced by Reflection AI. Vercel says reflection/beam transfers to paid pricing without downtime after the offer, while reflection/beam-free stops serving.
Mechanism: The two identifiers separate continuity through a pricing change from a hard stop when the free offer ends, making cost behavior a configuration choice. Why now: Beam’s newly announced Gateway availability makes the slug choice relevant to teams testing coding or agent workflows through a shared model API.
Implication: Teams should select the identifier that matches their budget and fallback expectations before integrating it into a workflow. Evidence boundary: The announcement gives no free-period end date and no independent benchmark or measured production performance results. Watch: Watch whether Vercel or Reflection AI announces an end date or pricing details, and whether teams migrate from the free slug before service stops.
This is a discrete availability and access-policy change, not evidence that Beam outperforms other models in production. The source documents the offer and its slug behavior; it does not report comparative task results.