Portkey vs Vercel AI Gateway: Which AI Gateway in 2026?
By API7.ai Team
Last updated: June 2026
Portkey and Vercel AI Gateway both put one API in front of many LLM providers, but take opposite paths: Portkey is an open-source gateway plus a hosted governance platform, while Vercel AI Gateway is a fully managed SaaS. This guide compares them on architecture, routing, caching, guardrails, observability, governance, self-hosting, and pricing so you can choose the right fit.
Portkey pairs an MIT TypeScript gateway with a hosted control plane — widest catalog, semantic caching, native guardrails, an MCP gateway, and a self-hostable OSS gateway (advanced governance is paywalled by tier). Vercel AI Gateway is a fully managed, zero-ops SaaS with hundreds of models and the best Vercel/Next.js DX, but it is not self-hostable and has no native guardrails or semantic cache. The core axis is an OSS gateway plus hosted platform versus a fully managed SaaS.
- Teams wanting an OSS gateway + hosted governance: Portkey
- Teams on Vercel/Next.js wanting zero-ops: Vercel AI Gateway
Portkey vs Vercel AI Gateway at a glance
Portkey leads on catalog breadth, native guardrails, semantic caching, and self-hosting; Vercel AI Gateway leads on zero-ops and Vercel/Next.js developer experience. Neither offers semantic routing or ensemble.
| Dimension | Portkey | Vercel |
|---|---|---|
| Best for | OSS gateway + hosted governance | Zero-ops on Vercel/Next.js |
| Core & form factor | TypeScript gateway + hosted control plane | Fully managed SaaS (proprietary) |
| Open-source license | MIT gateway; platform commercial | — Proprietary |
| Provider coverage | 1,600+ models / 45+ providers | Hundreds of models / ~45 providers |
| Semantic routing | — Rule-based only | — Not documented |
| Caching | ✓ Simple + semantic | — Prompt/automatic only |
| Guardrails | ✓ Native (tiered) | — None native |
| MCP gateway | ✓ Documented | — Platform/SDK, not the gateway |
| Self-host / VPC | OSS self-host; VPC = Enterprise | — Not self-hostable |
| SSO / SCIM | Enterprise-only | Enterprise-only |
What is Portkey?
Portkey is an AI gateway pairing an open-source TypeScript gateway with a paid hosted control plane for observability, prompt management, and governance.
Portkey is an AI gateway that pairs an open-source TypeScript gateway (MIT) with a hosted control plane for observability, prompt management, and governance. It advertises 1,600+ models across 45+ providers.
Language
TypeScript (Node.js)
License
MIT gateway + commercial platform
Form factor
OSS gateway + hosted SaaS
Best for
Hosted observability & governance
Pros
- 1,600+ models across 45+ providers
- Simple and semantic caching (semantic is Enterprise)
- Native guardrails and a documented MCP gateway
- OSS gateway is self-hostable via Docker/Kubernetes
Cons
- RBAC (Production tier), SSO/SCIM, VPC deploy, granular budgets are paid
- Conditional routing is rule/parameter-based, not semantic
- No documented ensemble; log/metric retention is tiered by plan
What is Vercel AI Gateway?
Vercel AI Gateway is a proprietary, fully managed SaaS that puts one API and one key in front of hundreds of models; it routes traffic through Vercel and is not self-hostable.
Vercel AI Gateway is a proprietary, fully managed SaaS that puts one API and one key in front of hundreds of models across roughly 45 providers. Traffic routes through Vercel; it is not self-hostable and is tuned for the Vercel and Next.js developer experience.
Language
Proprietary (managed service)
License
Proprietary
Form factor
Fully managed SaaS
Best for
Zero-ops on Vercel/Next.js
Pros
- Fully managed and zero-ops — no infrastructure to run
- Hundreds of models across ~45 providers behind one key
- Tightest Vercel/Next.js developer experience
- Built-in usage/spend and budgets; BYOK with no token markup
Cons
- Not self-hostable; no in-VPC or data-residency option
- No native guardrails and no semantic cache
- MCP is a Vercel platform/SDK feature, not part of the gateway
Portkey vs Vercel AI Gateway: feature comparison
The two converge on multi-provider access and spend controls, then diverge on form factor (OSS gateway + hosted platform vs fully managed SaaS), guardrails, caching, and self-hosting.
| Feature | Portkey | Vercel |
|---|---|---|
| Core & form factor | Open-source TypeScript gateway (Node.js) + a separate hosted control plane | Proprietary, fully managed SaaS; traffic routes through Vercel (ai-gateway.vercel.sh/v1) |
| Provider coverage | 1,600+ models across 45+ providers | Hundreds of models across ~45 providers, one key |
| Routing | Load balancing, fallbacks, retries, conditional routing (metadata/params/path), circuit breaker | Provider ordering/filtering (sort by cost/latency/throughput), per-provider timeouts, model fallbacks, auto-retry across providers |
| Semantic routing | — Rule/parameter-based only | — Not documented |
| Ensemble / fusion | — Not documented | — Not documented |
| Caching | Simple + semantic caching (semantic is Enterprise) | Prompt/automatic caching only — no semantic cache |
| Guardrails | 20+ deterministic native + LLM-based + partner checks; tiered Basic/Partner/Pro | No native guardrails; moderation model callable, Bedrock guardrails pass through |
| Observability & spend | Logging, tracing, 21+ metric dashboard, OpenTelemetry; retention tiered by plan | Built-in usage/spend dashboard and budgets; BYOK with no token markup |
| Governance | Model Catalog cost/token budgets, rate limits, workspaces; RBAC from Production tier | Usage/spend limits and budgets; identity governance via Vercel platform |
| MCP gateway | ✓ Documented (auth + access control for remote MCP) | — A Vercel platform/SDK feature, not the gateway |
| Self-host / VPC | OSS gateway self-hosts (Docker/K8s); managed VPC/hybrid is Enterprise | Not self-hostable; no in-VPC or data-residency option |
| Enterprise identity | RBAC from Production tier; SSO/SAML/OIDC & SCIM Enterprise | SSO/SCIM via Vercel Enterprise (SAML SSO also a Pro add-on) |
Pricing comparison
Portkey has a free open-source gateway and a tiered hosted platform; Vercel AI Gateway is a managed service billed through Vercel with BYOK and no token markup.
Portkey's gateway is free (MIT); the hosted platform is tiered — a free Developer tier with capped logs, a paid Production tier (from about $49/month) that unlocks RBAC and service accounts, and an Enterprise tier for SSO/SCIM, VPC/hybrid deployment, semantic caching, granular budgets, and custom retention. Vercel AI Gateway is a fully managed service billed through your Vercel account: you bring your own provider keys (BYOK) with no token markup, and use built-in usage/spend tracking and budgets to control cost. SSO and SCIM come with Vercel Enterprise (SAML SSO is also a Pro add-on). In short, Portkey gates breadth-of-platform features by tier, while Vercel folds the gateway into its platform pricing and identity model.
When to use Portkey vs Vercel AI Gateway
Choose Portkey for catalog breadth, native guardrails, and a self-hostable OSS gateway; choose Vercel AI Gateway for zero-ops and the tightest Vercel/Next.js experience.
Choose Portkey if you…
- Want the widest provider catalog behind one API
- Need native guardrails, semantic caching, or an MCP gateway
- Want to self-host the open-source gateway in your own infrastructure
Choose Vercel AI Gateway if you…
- Already build on Vercel or Next.js and want zero-ops
- Want a fully managed service with no infrastructure to run
- Want BYOK with no token markup and built-in spend controls
Bottom line
Choose Portkey for the widest catalog, native guardrails, semantic caching, and a self-hostable OSS gateway; choose Vercel AI Gateway for fully managed, zero-ops access that shines on Vercel and Next.js.
For the broadest catalog, native guardrails, semantic caching, an MCP gateway, and a self-hostable open-source gateway, Portkey is the stronger pick; for fully managed, zero-ops access that shines on Vercel and Next.js — with BYOK and no token markup — Vercel AI Gateway fits better. If you want a fully open, self-hosted data plane you run in your own VPC, AISIX is another option worth a look: a Rust, Apache-2.0 gateway from the creators of Apache APISIX, with semantic routing and ensemble built into the open-source core. See all AI gateway comparisons and AISIX vs LiteLLM.
Frequently asked questions
Related comparisons
Portkey vs LiteLLM · AISIX vs LiteLLM · All AI gateway comparisons
Ready to get started?
For more information about full API lifecycle management, please contact us to Meet with our API Experts.
