effect-vercel

Effect (v4) integrations for Vercel. Each one is its own entry point:

| Import | What | | ----------------------------- | ------------------------------------------------------------------------------- | | effect-vercel/ai-gateway | Vercel AI Gateway for Effect AI | | effect-vercel/oidc | The OIDC token Vercel issues to a deployment | | effect-vercel/runtime-cache | Runtime Cache for Effect Persistence |

Install

pnpm add effect-vercel effect
pnpm add @effect/ai-anthropic   # only for effect-vercel/ai-gateway
pnpm add @vercel/functions      # only for effect-vercel/runtime-cache

AI Gateway

The gateway serves the Anthropic Messages API and routes to any model it lists, named provider/model. AiGateway points Effect AI's Anthropic provider at the gateway, handles authentication per request, and smooths over the small differences between the gateway's dialect and what the provider's schemas expect.

Usage

import { AiGateway } from "effect-vercel/ai-gateway"
import { AnthropicLanguageModel } from "@effect/ai-anthropic"
import { Effect, Layer } from "effect"
import { LanguageModel } from "effect/ai"
import { FetchHttpClient } from "effect/http"

const Model = AnthropicLanguageModel.layer({ model: "google/gemini-2.5-flash" }).pipe(
  Layer.provide(AiGateway.layer),
  Layer.provide(FetchHttpClient.layer),
)

const program = LanguageModel.generateText({ prompt: "Say hi" }).pipe(
  Effect.map((r) => r.text),
  Effect.provide(Model),
)

AiGateway.layer provides AnthropicClient and requires an HttpClient. It authenticates with AI_GATEWAY_API_KEY if set, else the Vercel OIDC token, the same order as Vercel's own SDK. So it works locally with an API key and on Vercel with no setup.

Any Effect AI feature that works with the Anthropic provider (tools, structured output, streaming, thinking) works through the gateway, for every model the gateway offers.

Credentials

To pick the source yourself, use AiGateway.layerWithoutCredentials and provide an AiGatewayCredentials layer:

AnthropicLanguageModel.layer({ model: "google/gemini-2.5-flash" }).pipe(
  Layer.provide(AiGateway.layerWithoutCredentials),
  Layer.provide([AiGatewayCredentials.layerFromApiKey(key), FetchHttpClient.layer]),
)

| Layer | Source | | ------------------------------------------- | --------------------------------------------------------- | | AiGatewayCredentials.layer | The default: layerFromEnv, then layerFromVercelOidc. | | AiGatewayCredentials.layerFromEnv | AI_GATEWAY_API_KEY. Local dev, CI, anywhere off Vercel. | | AiGatewayCredentials.layerFromVercelOidc | The Vercel OIDC token (see OIDC). | | AiGatewayCredentials.layerFromApiKey(key) | A fixed key. Tests, custom wiring. |

The credential is resolved on every request, so a rotating token is always current. Each request carries x-api-key and ai-gateway-auth-method (api-key or oidc). A missing credential fails the request with an AiError wrapping an AiGatewayCredentialsError that names the source it tried and how to fix it.

Environment variables are read through Effect's Config, so a ConfigProvider can redirect them.

Dialect fixes

Two adjustments are applied to traffic with the gateway:

• Requests: "cache_control": null is removed from content blocks. The Effect provider emits it, Anthropic accepts it, the gateway rejects it.
• Responses: keys Anthropic always sends but the gateway omits are filled in with Anthropic's "nothing to report" values, so the provider's schemas decode. These are the cache and service-tier usage fields, signature on thinking blocks from non-Anthropic models, and type/request_id on error envelopes (unknown error.type values map to api_error).

Streaming responses pass through untouched.

OIDC

Vercel issues each deployment an OIDC token. Use it for the AI Gateway, or exchange it for cloud credentials (AWS AssumeRoleWithWebIdentity, GCP workload identity federation).

import { VercelOidc } from "effect-vercel/oidc"
import { Effect, Redacted } from "effect"

const program = Effect.gen(function* () {
  const token = yield* VercelOidc.token
  // Redacted.value(token) is the web identity token
})

• VercelOidc.token reads the x-vercel-oidc-token header of the current request (Functions), else VERCEL_OIDC_TOKEN (builds, vercel env pull). It fails with a VercelOidcError when neither is set.
• VercelOidc.find is the same but returns Option.none() instead of failing, for fallback chains.

Both read the token on every run, since it rotates per request. They do not refresh an expired local token; re-run vercel env pull.

Runtime Cache

VercelRuntimeCache.layer provides Effect's Persistence, stored in the Vercel Runtime Cache. Use it with PersistedCache to share results across Function instances in a region.

import { VercelRuntimeCache } from "effect-vercel/runtime-cache"
import { Effect, Schema } from "effect"
import { Persistable, PersistedCache } from "effect/persistence"

class GetUser extends Persistable.Class<{ payload: { id: number } }>()("GetUser", {
  primaryKey: (req) => `GetUser:${req.id}`,
  success: User,
  error: Schema.Never,
}) {}

const program = Effect.gen(function* () {
  const users = yield* PersistedCache.make(fetchUser, {
    storeId: "users",
    timeToLive: () => "1 hour",
  })
  return yield* users.get(new GetUser({ id: 1 }))
}).pipe(Effect.scoped, Effect.provide(VercelRuntimeCache.layer))

• Off Vercel, @vercel/functions falls back to an in-memory cache, so the layer works in local dev and tests.
• Entries are keyed by [storeId, primaryKey], hashed with SHA-256. Only the hash reaches Vercel. getCache's default hash is 32-bit and can collide, so VercelRuntimeCache.makeBacking(cache) expects a cache with a stronger keyHashFunction if you build your own.
• Each entry is tagged effect-persistence:<sha256 of storeId>. clear expires that tag.
• TTLs round up to whole seconds.
• The Runtime Cache swallows its own network errors and timeouts, which then read as misses and silently skipped writes, removes and clears. getCache has no option to surface them.

Development

pnpm install
pnpm test
pnpm check   # tsc
pnpm lint    # oxlint
pnpm fmt     # oxfmt
pnpm build   # tsdown -> dist/

License

MIT