Is Conifer an alternative to Vercel AI Gateway?

Yes. Both are hosted gateways that give one API key for models from many providers, with fallbacks, budgets and the cost of each request. Vercel AI Gateway also serves image, video, speech and realtime models and works with Vercel’s AI SDK. Conifer serves a named model as named or refuses it, and refuses a request over a cost ceiling before any provider is called.

How do I move from Vercel AI Gateway to Conifer?

Change the base URL to Conifer’s and use a Conifer key. Model names with the lab in front are accepted, so they need no rewriting. A model Conifer does not carry is refused rather than replaced.

Do Vercel AI Gateway and Conifer add a fee to model prices?

Vercel AI Gateway charges the provider’s list price with no markup or platform fee on tokens. Conifer bills every model at its maker’s published list price per token, with no subscription, and adds no fee on requests served by your own Azure OpenAI resource. On both, you pay the card processing fee on a top-up.

Does Conifer store my prompts?

The gateway does not store the prompt of a standard request. It holds a response for ten minutes so that a retried request is not charged twice, and keeps request details such as the model, token counts and cost for billing.

Comparison

Conifer and Vercel AI Gateway compared

Both are hosted gateways that give one API key for AI models from many providers. This page sets out how Conifer and Vercel AI Gateway each work, using each company’s own published documentation.

Summary

Vercel AI Gateway is a managed gateway run by Vercel. Apps can call it from any server, whether or not they run on Vercel, and it covers text, image, video, speech, transcription, realtime, embedding and reranking models at the provider’s list price, with no markup on tokens. Conifer is a hosted gateway for more than 230 models that serves a named model as named or refuses it, states the exact cost on every response, and refuses a request over a cost ceiling before any provider is called.

Vercel AI Gateway details were read from vercel.com and its documentation on 7 November 2026. Plans and prices change; check Vercel’s pages before you decide.

Side by side

Topic	Conifer	Vercel AI Gateway
Where it runs	A hosted service at one Conifer address; there is nothing to deploy	A managed gateway run by Vercel. Apps call it with an API key from any server, cloud or local machine, and apps deployed on Vercel can use short-lived tokens instead
Models	More than 230 models from Anthropic, OpenAI, Google and other labs, paid from one prepaid balance	“400+ models” for text, image, video, audio, realtime, retrieval and decisions. Free credits cover a subset of them
Interfaces	OpenAI Chat Completions, OpenAI Responses and Anthropic Messages on one key, plus embeddings	Vercel’s AI SDK, OpenAI Chat Completions, OpenAI Responses, Anthropic Messages and OpenResponses, plus its own REST endpoints
What you pay	Every model is billed at its maker’s published list price per token, from a prepaid balance with no subscription. A top-up adds a small processing line that covers Stripe’s card fee	The provider’s list price, with no markup and no platform fee on tokens, paid from prepaid credits. A monthly free credit covers a subset of models. You pay any payment processing fees, Enterprise teams can pay by invoice, and some options cost extra
Your own provider keys	Your own Azure OpenAI resource can sit behind the same key, and requests it serves carry no Conifer fee. AWS Bedrock and Google Vertex are not served as your own endpoints today	Your own key for any provider in its catalog, with no markup or fee, once you have bought credits. If a request with your key fails, it is retried with Vercel’s credentials and charged to your credits
A named model	Served as named, or refused. Another model is tried if you list it as a fallback yourself, up to three	Fallback models you list are tried in order if the first fails. A team routing rule can serve requests for one model with another model, or block a model
Provider failover	When a provider fails, rate-limits or times out, the gateway serves the same model from another provider at the same price, for models with more than one provider	Providers are chosen by recent uptime and latency, and a failed request is retried on the model’s other providers. You can set the provider order or limit the providers
Cost on each request	Every response carries the amount charged, itemised by token type, and a streamed response ends with the same receipt	Request logs show the cost of every request, and a REST endpoint returns the cost of a single request
Spending limits	A cost ceiling per request, a spend cap per key and a monthly limit on the account. A request that could exceed a ceiling is refused before any provider is called	Budgets for a team, project, API key or team member, reset daily, weekly or monthly. A budget is a soft cap: the request that crosses it still completes. Spend on your own provider keys is not counted
Data handling	The prompt of a standard request is not stored. A response is held for ten minutes so a retry is not charged twice, and request details such as model, tokens and cost are kept for billing	Prompts and responses are deleted once the request completes, and details such as model, provider, tokens and cost are logged. Routing to providers with zero data retention is free per request and $0.10 per 1,000 requests team-wide
Other model types	Chat models and embeddings. Image, audio, Files and Batches endpoints have no Conifer equivalent today, and the moderations endpoint is not served	Image, video, speech, transcription, realtime, embedding and reranking models through the same gateway
Enterprise terms	Team accounts with owner, admin and member roles, per-key model scope and spend caps, set up on request	Invoice payment and volume discounts for Enterprise, purchase through AWS Marketplace, and a team-wide provider allowlist on Pro and Enterprise

Which one fits

Vercel AI Gateway may suit you if: You already build with Vercel or its AI SDK, you need image, video, speech or realtime models behind the same key, you want to bring your own keys for many providers, or you want to buy through AWS Marketplace or pay by invoice.
Conifer may suit you if: You need each response to state the model that served and the exact amount charged, you want a cost ceiling on each request enforced before any provider is called, you want a named model served as named unless you list a fallback, or you want a router that chooses from a pool of up to 64 models you approve.

Moving from Vercel AI Gateway to Conifer

Address and key: Point the OpenAI or Anthropic client you use with Vercel AI Gateway at Conifer, and use a Conifer key.
baseURL: "https://api.conifer.build/v1", apiKey: process.env.CONIFER_API_KEY,
Model names: Names with the lab in front, such as anthropic/claude-opus-5, are accepted and served as the matching model in the Conifer catalog. A model Conifer does not carry is refused rather than replaced.
Fallbacks: Conifer tries another provider for the same model on its own. To try other models, list up to three fallbacks in a request header.
Provider choice: Conifer picks the provider for each request. Setting your own provider order has no Conifer equivalent.
Check first: Image, audio, Files and Batches endpoints have no Conifer equivalent, and the moderations endpoint is not served. Check each model’s capabilities on the models page before you move.

Coming from another gateway shows the change for Vercel AI Gateway, Replace your provider key has the full steps, and Routing covers fallbacks and failover.

Conifer Gateway: One API key and endpoint for every model in the catalog, with the model and cost stated on every response.
What is an LLM gateway?: How a gateway sits between your software and the companies that serve models, and what to check when you choose one.
Conifer vs OpenRouter: The two gateways side by side, from each company's own documentation.
Conifer vs Portkey: The two gateways side by side, from each company's own documentation.
Conifer vs Helicone: The two gateways side by side, from each company's own documentation.
All comparisons: Conifer compared with eleven other LLM gateways and routers, from their own documentation.

Questions

Is Conifer an alternative to Vercel AI Gateway?: Yes. Both are hosted gateways that give one API key for models from many providers, with fallbacks, budgets and the cost of each request. Vercel AI Gateway also serves image, video, speech and realtime models and works with Vercel’s AI SDK. Conifer serves a named model as named or refuses it, and refuses a request over a cost ceiling before any provider is called.
How do I move from Vercel AI Gateway to Conifer?: Change the base URL to Conifer’s and use a Conifer key. Model names with the lab in front are accepted, so they need no rewriting. A model Conifer does not carry is refused rather than replaced.
Do Vercel AI Gateway and Conifer add a fee to model prices?: Vercel AI Gateway charges the provider’s list price with no markup or platform fee on tokens. Conifer bills every model at its maker’s published list price per token, with no subscription, and adds no fee on requests served by your own Azure OpenAI resource. On both, you pay the card processing fee on a top-up.
Does Conifer store my prompts?: The gateway does not store the prompt of a standard request. It holds a response for ten minutes so that a retried request is not charged twice, and keeps request details such as the model, token counts and cost for billing.