# Virtual models now available on AI Gateway

Source: [Vercel](https://vercel.com/changelog/virtual-models-now-available-on-ai-gateway)  
Feed7 permalink: https://feed7.dev/p/virtual-models-now-available-on-ai-gateway-1nnw9v7  
Published: 2026-10-28T00:00:00.000Z  
Trust: Official Source (official_source)

## Why Included

AI Gateway Virtual Models now let apps and coding agents use a task ID for centrally managed model settings, routing, and spend tags. They can also set options an agent may not expose directly.

## Source Summary

AI Gateway Virtual Models are saved configurations addressed by IDs such as **vmc/coding-agent**. Teams can centrally set the model, reasoning effort, fast mode, prompt caching, provider routing, fallbacks, Zero Data Retention, and reporting tags.

## Practical Implication

Builders can point compatible API requests or coding agents at a virtual model to manage these settings without reconfiguring each caller. This is useful when agent harnesses expose only a subset of provider options.

## Agent-Ready Context

AI Gateway Virtual Models are saved configurations addressed by IDs such as **vmc/coding-agent**. Teams can centrally set the model, reasoning effort, fast mode, prompt caching, provider routing, fallbacks, Zero Data Retention, and reporting tags.

Builders can point compatible API requests or coding agents at a virtual model to manage these settings without reconfiguring each caller. This is useful when agent harnesses expose only a subset of provider options.

The changelog says virtual models work with every model type on AI Gateway, but does not detail which settings each agent harness supports or how configuration changes are tested.

## Connected Context

Feed7 judgment across 955 accumulated Signals:

This changelog confirms the stable-ID approach as a saved, centrally managed configuration spanning more routing and compliance controls than the blog’s shorter overview lists. It strengthens the practical case for configuring agent behavior outside individual callers, while leaving harness support and configuration testing unspecified.

- [Introducing Virtual Models on AI Gateway](https://feed7.dev/p/ai-gateway-virtual-models-0563pz8) — The blog introduces the same stable task-ID mechanism and advises evaluating a candidate before production changes; this changelog confirms the feature and names additional centrally managed settings, without adding an evaluation method.
- [AI Gateway adds unified fast mode support](https://feed7.dev/p/ai-gateway-adds-unified-fast-mode-support-144dq26) — Unified fast mode is one of the routing controls that Virtual Models can centrally configure, making it possible to apply latency preferences through a task ID; fast-mode coverage and cost still need to be considered.
- [Claude Opus 5 now available on AI Gateway](https://feed7.dev/p/claude-opus-5-now-available-on-ai-gateway-16oaf27) — Opus 5 exposes reasoning, fast mode, fallback, and ZDR options that fit the central configuration model described here. Its safeguards and lack of comparative outcome evidence still make task-specific evaluation relevant.

## Context Map

- Layer: infra
- Domains: coding, data
- Topics: gateways, model-selection, coding-agents

## Uncertainty

- The changelog says virtual models work with every model type on AI Gateway, but does not detail which settings each agent harness supports or how configuration changes are tested.

## Agent Instruction

Use this item as source-backed context. Do not invent claims beyond the linked source. If this item conflicts with another source, call out the conflict.