{"slug": "service-tiers-now-available-on-ai-gateway", "title": "Service tiers now available on AI Gateway", "summary": "Vercel's AI Gateway now supports service tiering, allowing users to optimize for latency, throughput, and cost per request across OpenAI and Gemini models. The feature offers three tiers: default (standard), priority (faster at ~1.8-2x cost), and flex (lower cost at ~0.5x with higher latency), with billing adjusted based on the tier actually used.", "body_md": "AI Gateway now supports service tiering. Service tiers let you optimize for latency, throughput, and cost per request to match your use case. Pick a faster tier for interactive workloads (less queueing, higher token throughput), or a lower cost tier for background jobs that can tolerate more latency.\n\nAt launch, service tiering is available for OpenAI and Gemini models.\n\n[Service tiers](https://vercel.com/docs/ai-gateway/models-and-providers/service-tiers) work across every AI Gateway API format: [AI SDK](https://ai-sdk.dev), [Chat Completions API](https://vercel.com/docs/ai-gateway/sdks-and-apis/openai-chat-completions), [Anthropic Messages API](https://vercel.com/docs/ai-gateway/sdks-and-apis/anthropic-messages-api), [OpenAI Responses API](https://vercel.com/docs/ai-gateway/sdks-and-apis/responses), and [OpenResponses API](https://vercel.com/docs/ai-gateway/sdks-and-apis/openresponses). AI Gateway adjusts billing based on the tier each request used.\n\n`default`\n\n: Standard processing\n\n`priority`\n\n: Faster processing at increased cost\n\n`flex`\n\n: Lower cost with potentially higher latency\n\nIf a service tier is not specified, requests run on the default tier.\n\nSet `serviceTier`\n\nunder `providerOptions.gateway`\n\n. The same option works across all models and providers that support service tiers, so you can swap the model without restructuring provider-specific options:\n\nThe applied tier is returned in the provider metadata so you can confirm which tier served the request. This is useful when a priority request falls back to default capacity, or when comparing observed latency across tiers.\n\nService tier is serviced on a best-effort basis: if a tier can't be applied, the request runs on the default tier at the default rate, and only an invalid service tier value fails the request.\n\nIf a model can be served by multiple providers and you only want a service tier on some of them or to configure different service tiers across each, set the tier on the provider's own namespace instead.\n\nAI Gateway applies per-tier rates automatically based on the tier each request actually used. If a `priority`\n\nrequest gets downgraded to default capacity, billing reflects the default rate, not the priority rate.\n\n|\n|\n|\n|\n| Baseline | ~1.8-2x default pricing | ~0.5x default pricing |\n\nMore details about providers and tiering are in the [service tiers reference](https://vercel.com/docs/ai-gateway/models-and-providers/service-tiers).", "url": "https://wpnews.pro/news/service-tiers-now-available-on-ai-gateway", "canonical_source": "https://vercel.com/changelog/service-tiers-now-available-on-ai-gateway", "published_at": "2026-07-21 00:00:00+00:00", "updated_at": "2026-07-21 22:08:36.045706+00:00", "lang": "en", "topics": ["ai-infrastructure", "ai-products", "developer-tools"], "entities": ["Vercel", "AI Gateway", "OpenAI", "Gemini"], "alternates": {"html": "https://wpnews.pro/news/service-tiers-now-available-on-ai-gateway", "markdown": "https://wpnews.pro/news/service-tiers-now-available-on-ai-gateway.md", "text": "https://wpnews.pro/news/service-tiers-now-available-on-ai-gateway.txt", "jsonld": "https://wpnews.pro/news/service-tiers-now-available-on-ai-gateway.jsonld"}}