gridaco · softmarshmallow · Jun 6, 2026 · Jun 6, 2026
diff --git a/docs/contributing/billing.md b/docs/contributing/billing.md
@@ -26,7 +26,7 @@ BYOK_OPENROUTER_API_KEY=sk-or-v1-...      # https://openrouter.ai/keys
 
 - Bypasses **billing only — never auth.** Still sign in (`insider@grida.co` / `password`); a resolvable org is still required (an unauthenticated request still 401s).
 - **Text/chat only** — BYOK swaps the AI-SDK provider, so only the text path is unbilled. Image/audio go through Replicate (`withTransaction`) and **still gate + bill even under BYOK** — those features need the full billing setup. (OpenRouter also exposes no image/audio models.) Catalog model IDs are unchanged; use IDs your provider accepts (edit `editor/lib/ai/models.ts` locally if one 404s).
-- Precedence if both are set: OpenRouter, then AI Gateway. Fail-closed — an empty/unset (or whitespace-only) key falls back to the billed path.
+- Precedence if both are set: OpenRouter, then Vercel. Fail-closed — an empty/unset (or whitespace-only) key falls back to the billed path.
 - **Never set `BYOK_*` on a hosted or preview deploy.** It disables billing **and** the org-id sanity gate for every org. Contributor / self-host / local only. See [SECURITY.md](https://github.com/gridaco/grida/blob/main/SECURITY.md) (`GRIDA-SEC-003`, BYOK carve-out).
 
 Working on billing itself? Ignore BYOK and continue with the full setup below.

diff --git a/docs/wg/ai/grida/architecture.md b/docs/wg/ai/grida/architecture.md
@@ -122,7 +122,7 @@ There is intentionally no `/secrets/get`, no `/auth/*`, and no
 V1 provider resolution is BYOK-only:
 
 1. `openrouter`
-2. `ai-gateway`
+2. `vercel`
 3. unavailable (`provider_down`)
 
 `AgentRunOptions.providerId` accepts only `ByokProviderId`. The package root
@@ -215,8 +215,8 @@ describe("handshake", () => {
 });
 
 describe("provider resolution", () => {
-  it("prefers OpenRouter BYOK over AI Gateway BYOK");
-  it("falls back to AI Gateway BYOK");
+  it("prefers OpenRouter BYOK over Vercel BYOK");
+  it("falls back to Vercel BYOK");
   it("throws provider_down when no BYOK key is present");
   it("validates explicit BYOK provider ids");
 });

diff --git a/editor/.env.example b/editor/.env.example
@@ -38,7 +38,7 @@ OPENAI_API_KEY='sk-xxx'
 # route through that provider and BYPASS the billing layer (no credit
 # gate, no metering). does NOT bypass auth (login + org still required).
 # text/chat only; catalog model IDs unchanged (use IDs the provider
-# accepts). precedence: openrouter, then ai-gateway.
+# accepts). precedence: openrouter, then vercel.
 # BYOK_OPENROUTER_API_KEY=sk-or-v1-xxx   # https://openrouter.ai/keys
 # BYOK_AI_GATEWAY_API_KEY=...            # dedicated Vercel AI Gateway key
 

diff --git a/editor/lib/ai/models.ts b/editor/lib/ai/models.ts
@@ -101,7 +101,7 @@ export const gateway = createGateway({
 // active only when a key env var is a non-empty string after trim
 // (whitespace-only secrets fall back to the billed path).
 //
-// Implementations (precedence: OpenRouter first, then AI Gateway). A
+// Implementations (precedence: OpenRouter first, then Vercel). A
 // third BYOK key is a new branch here — no registry.
 // ---------------------------------------------------------------------------
 function resolveByokProvider() {

diff --git a/editor/lib/desktop/bridge.test.ts b/editor/lib/desktop/bridge.test.ts
@@ -45,7 +45,7 @@ describe("desktop bridge client contract", () => {
 
   it("uses producer-owned BYOK provider metadata for settings", () => {
     expect(secrets.byokProviderMetadata()).toBe(BYOK_PROVIDER_METADATA);
-    expect(secrets.byokProviders()).toEqual(["openrouter", "ai-gateway"]);
+    expect(secrets.byokProviders()).toEqual(["openrouter", "vercel"]);
   });
 
   it("rejects empty keys before calling the bridge", async () => {

diff --git a/packages/grida-ai-agent/src/__public-api__.test.ts b/packages/grida-ai-agent/src/__public-api__.test.ts
@@ -100,14 +100,14 @@ describe("@grida/agent public API", () => {
     });
 
     it("exposes BYOK provider identity, wire vocab, tiers, and session-row types", () => {
-      expect(BYOK_PROVIDER_IDS).toEqual(["openrouter", "ai-gateway"]);
+      expect(BYOK_PROVIDER_IDS).toEqual(["openrouter", "vercel"]);
       expect(BYOK_PROVIDER_METADATA.map((provider) => provider.label)).toEqual([
         "OpenRouter",
-        "AI Gateway",
+        "Vercel",
       ]);
-      const byok: ByokProviderId = "ai-gateway";
+      const byok: ByokProviderId = "vercel";
       const metadata: ByokProviderMetadata = BYOK_PROVIDER_METADATA[0];
-      expect(byok).toBe("ai-gateway");
+      expect(byok).toBe("vercel");
       expect(metadata.id).toBe("openrouter");
 
       // Tier constants.

diff --git a/packages/grida-ai-agent/src/auth/file.ts b/packages/grida-ai-agent/src/auth/file.ts
@@ -11,8 +11,8 @@
  *
  *   ```json
  *   {
- *     "openrouter":  { "type": "api", "key": "sk-or-..." },
- *     "ai-gateway":  { "type": "api", "key": "..." }
+ *     "openrouter": { "type": "api", "key": "sk-or-..." },
+ *     "vercel":     { "type": "api", "key": "..." }
  *   }
  *   ```
  *
@@ -97,7 +97,7 @@ export class AuthStore {
    * Serializes `set` / `remove` so two concurrent mutations can't
    * lose-update each other. Both ops follow read-modify-write on a
    * shared file; without this chain a parallel set('openrouter', …)
-   * and set('ai-gateway', …) would both `readAll` the same starting
+   * and set('vercel', …) would both `readAll` the same starting
    * state, both `writeAll`, and the second `rename` wins — silently
    * dropping the first key. The `.catch(() => undefined)` on the
    * chain swallows rejection so a single failed write doesn't strand

diff --git a/packages/grida-ai-agent/src/http/routes/secrets.ts b/packages/grida-ai-agent/src/http/routes/secrets.ts
@@ -12,7 +12,7 @@
  *
  * Allowed provider ids — a closed set:
  *   - `openrouter`
- *   - `ai-gateway`
+ *   - `vercel`
  *
  * Any other id is rejected with a 400 so a typo doesn't silently create a
  * never-used auth.json entry.

diff --git a/packages/grida-ai-agent/src/protocol/provider-ids.ts b/packages/grida-ai-agent/src/protocol/provider-ids.ts
@@ -11,8 +11,8 @@ export const BYOK_PROVIDER_METADATA = [
     label: "OpenRouter",
   },
   {
-    id: "ai-gateway",
-    label: "AI Gateway",
+    id: "vercel",
+    label: "Vercel",
   },
 ] as const;
 

diff --git a/packages/grida-ai-agent/src/providers/byok.ts b/packages/grida-ai-agent/src/providers/byok.ts
@@ -32,7 +32,7 @@ export function makeOpenRouterFactory(apiKey: string): ModelFactory {
   return (tier, modelId) => provider(modelId ?? MODEL_BY_TIER[tier]);
 }
 
-export function makeAiGatewayFactory(apiKey: string): ModelFactory {
+export function makeVercelFactory(apiKey: string): ModelFactory {
   const provider = createGateway({ apiKey });
   return (tier, modelId) => provider(modelId ?? MODEL_BY_TIER[tier]);
 }
diff --git a/packages/grida-ai-agent/src/providers/index.test.ts b/packages/grida-ai-agent/src/providers/index.test.ts
@@ -15,11 +15,11 @@ function deps(keys: Record<string, string | null> = {}) {
 }
 
 describe("resolveProvider", () => {
-  it("prefers OpenRouter over AI Gateway when both BYOK keys exist", async () => {
+  it("prefers OpenRouter over Vercel when both BYOK keys exist", async () => {
     const provider = await resolveProvider(
       deps({
         openrouter: " sk-or ",
-        "ai-gateway": "ai-gateway-key",
+        vercel: "vercel-key",
       })
     );
 
@@ -28,12 +28,10 @@ describe("resolveProvider", () => {
     expect(provider.model_factory).toBeTypeOf("function");
   });
 
-  it("falls back to AI Gateway when OpenRouter is absent", async () => {
-    const provider = await resolveProvider(
-      deps({ "ai-gateway": "ai-gateway-key" })
-    );
+  it("falls back to Vercel when OpenRouter is absent", async () => {
+    const provider = await resolveProvider(deps({ vercel: "vercel-key" }));
 
-    expect(provider.provider_id).toBe("ai-gateway");
+    expect(provider.provider_id).toBe("vercel");
     expect(provider.kind).toBe("byok");
   });
 
@@ -46,9 +44,9 @@ describe("resolveProvider", () => {
   it("throws when an explicit BYOK provider has no key", async () => {
     await expect(
       resolveProvider(deps({ openrouter: "sk-or" }), {
-        explicit: "ai-gateway",
+        explicit: "vercel",
       })
-    ).rejects.toMatchObject({ provider_id: "ai-gateway" });
+    ).rejects.toMatchObject({ provider_id: "vercel" });
   });
 
   it("BYOK factory honors an explicit modelId over the tier model", async () => {

diff --git a/packages/grida-ai-agent/src/providers/index.ts b/packages/grida-ai-agent/src/providers/index.ts
@@ -8,7 +8,7 @@
  * the factory, so it's cheap on the hot path and easy to test.
  *
  * This is the providers layer, not a generic model-provider router. V1 is
- * BYOK-only: OpenRouter takes precedence over AI Gateway, and a missing
+ * BYOK-only: OpenRouter takes precedence over Vercel, and a missing
  * key throws `ProviderUnavailableError`.
  */
 
@@ -20,7 +20,7 @@ import {
   BYOK_PROVIDER_METADATA,
   type ByokProviderId,
 } from "../protocol/provider-ids";
-import { makeAiGatewayFactory, makeOpenRouterFactory } from "./byok";
+import { makeOpenRouterFactory, makeVercelFactory } from "./byok";
 
 /** Canonical tier->catalog-model map. One table, sourced from @grida/ai-models. */
 export const MODEL_BY_TIER: Record<ModelTier, TierModelId> = TIER_MODEL_IDS;
@@ -98,11 +98,11 @@ function makeResolvedProvider(
         kind: "byok",
         model_factory: makeOpenRouterFactory(key.trim()),
       };
-    case "ai-gateway":
+    case "vercel":
       return {
         provider_id: providerId,
         kind: "byok",
-        model_factory: makeAiGatewayFactory(key.trim()),
+        model_factory: makeVercelFactory(key.trim()),
       };
   }
   const _exhaustive: never = providerId;

diff --git a/packages/grida-ai-agent/src/sandbox/policy.ts b/packages/grida-ai-agent/src/sandbox/policy.ts
@@ -40,7 +40,7 @@ export type AgentHostSandboxPolicy = {
 
 const BYOK_PROVIDER_NETWORK_HOSTS = {
   openrouter: ["openrouter.ai"],
-  "ai-gateway": ["ai-gateway.vercel.sh", "*.vercel-ai.com"],
+  vercel: ["ai-gateway.vercel.sh", "*.vercel-ai.com"],
 } as const satisfies Record<ByokProviderId, readonly string[]>;
 
 const ALWAYS_ALLOWED_HOSTS: readonly string[] = Object.values(

diff --git a/packages/grida-ai-agent/src/secrets.ts b/packages/grida-ai-agent/src/secrets.ts
@@ -2,7 +2,7 @@
  * GRIDA-SEC-004 — BYOK secret store.
  *
  * Thin façade over `AuthStore` for the `ApiKeyEntry`-shaped records
- * (`openrouter`, `ai-gateway`). The agent host's `/secrets/*` HTTP routes
+ * (`openrouter`, `vercel`). The agent host's `/secrets/*` HTTP routes
  * call this; the BYOK provider path in `runtime.ts`
  * calls this internally to pull the key when constructing the
  * @ai-sdk client.

diff --git a/packages/grida-ai-models/__tests__/models.test.ts b/packages/grida-ai-models/__tests__/models.test.ts
@@ -1,15 +1,15 @@
 import models, { TIER_MODEL_IDS } from "..";
 
 describe("models.image.findImageModelCard", () => {
-  it("resolves a full gateway id", () => {
+  it("resolves a full vercel id", () => {
     const card = models.image.findImageModelCard("bfl/flux-pro-1.1");
     expect(card?.id).toBe("bfl/flux-pro-1.1");
     expect(card?.label).toBe("Flux Pro 1.1");
   });
 
   it("resolves the deprecated ProviderModel wrapper", () => {
     const card = models.image.findImageModelCard({
-      provider: "gateway",
+      provider: "vercel",
       modelId: "bfl/flux-2-pro",
     });
     expect(card?.id).toBe("bfl/flux-2-pro");

diff --git a/packages/grida-ai-models/src/models.ts b/packages/grida-ai-models/src/models.ts
@@ -19,7 +19,7 @@
  * modules — keeping the full `namespace models` declaration in a
  * single source file is the workaround.
  *
- * `provider: "gateway"` and `provider: "replicate"` on the cards are
+ * `provider: "vercel"` and `provider: "replicate"` on the cards are
  * data labels only — see the README for the full contract.
  *
  * @module
@@ -31,11 +31,11 @@ export namespace models {
   // ── Shared discriminators ─────────────────────────────────────────
 
   /**
-   * Routing label for hosted-provider calls. `"gateway"` indicates
-   * the model is served via a hosted AI gateway (e.g. Vercel AI
-   * Gateway); the label is data, not an SDK directive.
+   * Routing label for hosted-provider calls. `"vercel"` indicates
+   * the model is served via the Vercel AI Gateway; the label is
+   * data, not an SDK directive.
    */
-  export type Provider = "gateway";
+  export type Provider = "vercel";
 
   /**
    * Model vendor (the organization that produced the weights).
@@ -226,7 +226,7 @@ export namespace models {
      * a `provider` field on its own.
      */
     export type ProviderModel = {
-      provider: "gateway";
+      provider: "vercel";
       modelId: ImageModelId;
     };
 
@@ -461,7 +461,7 @@ export namespace models {
         short_description:
           "State-of-the-art image generation and editing with flexible resolutions",
         vendor: "openai",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "medium",
         speed_max: "1m",
         styles: null,
@@ -514,7 +514,7 @@ export namespace models {
         short_description:
           "Previous-generation image model. Superseded by GPT Image 2.",
         vendor: "openai",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "medium",
         speed_max: "1m",
         styles: null,
@@ -561,7 +561,7 @@ export namespace models {
         deprecated: false,
         short_description: "Cost-efficient image generation model",
         vendor: "openai",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "slow",
         speed_max: "1m",
         styles: null,
@@ -613,7 +613,7 @@ export namespace models {
         short_description:
           "Fast, efficient multimodal model with native image generation",
         vendor: "google",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "fast",
         speed_max: "15s",
         styles: null,
@@ -636,7 +636,7 @@ export namespace models {
         short_description:
           "High-quality multimodal model with native image generation",
         vendor: "google",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "medium",
         speed_max: "30s",
         styles: null,
@@ -662,7 +662,7 @@ export namespace models {
         short_description:
           "Latest Flux model with best-in-class image quality and prompt adherence",
         vendor: "black-forest-labs",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "medium",
         speed_max: "30s",
         styles: null,
@@ -683,7 +683,7 @@ export namespace models {
         short_description:
           "Highest quality Flux model for context-aware image generation and editing",
         vendor: "black-forest-labs",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "slow",
         speed_max: "30s",
         styles: null,
@@ -703,7 +703,7 @@ export namespace models {
         deprecated: false,
         short_description: "Fast context-aware image generation and editing",
         vendor: "black-forest-labs",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "medium",
         speed_max: "20s",
         styles: null,
@@ -724,7 +724,7 @@ export namespace models {
         short_description:
           "Faster, better FLUX Pro. Text-to-image model with excellent image quality and output diversity.",
         vendor: "black-forest-labs",
-        provider: "gateway",
+        provider: "vercel",
         speed_label: "slow",
         speed_max: "30s",
         styles: null,