diff --git a/docs/.generated/plugin-sdk-api-baseline.sha256 b/docs/.generated/plugin-sdk-api-baseline.sha256
index 0ea0f6412a87..3e56bf69708e 100644
--- a/docs/.generated/plugin-sdk-api-baseline.sha256
+++ b/docs/.generated/plugin-sdk-api-baseline.sha256
@@ -1,2 +1,2 @@
-5cc5b1f35c16b3b65871942b262467b1eee065652f4f90a24ddd65eb1096aa3c  plugin-sdk-api-baseline.json
-aaeb88388cf472b788ba2332890811d5bee64208a99596c975b2f11f7d40c04e  plugin-sdk-api-baseline.jsonl
+38de898c2b05a0895c4863373aa7128f073128cc6ace8bd5004203c3ee9735fd  plugin-sdk-api-baseline.json
+93ca4d93917be8be643293c2a11c58d2a7c138390c15f646991a223815ddec1e  plugin-sdk-api-baseline.jsonl
diff --git a/packages/speech-core/api.ts b/packages/speech-core/api.ts
deleted file mode 100644
index 1696598b609f..000000000000
--- a/packages/speech-core/api.ts
+++ /dev/null
@@ -1,56 +0,0 @@
-// Public speech-core plugin SDK facade. Re-export stable provider/config helpers
-// from the plugin-sdk alias so speech plugins do not import core internals.
-export {
-  asBoolean,
-  asFiniteNumber,
-  asObject,
-  assertOkOrThrowProviderError,
-  canonicalizeSpeechProviderId,
-  createProviderHttpError,
-  extractProviderErrorDetail,
-  extractProviderRequestId,
-  formatProviderErrorPayload,
-  formatProviderHttpErrorMessage,
-  getSpeechProvider,
-  listSpeechProviders,
-  normalizeApplyTextNormalization,
-  normalizeLanguageCode,
-  normalizeSeed,
-  normalizeSpeechProviderId,
-  normalizeTtsAutoMode,
-  parseTtsDirectives,
-  readResponseTextLimited,
-  requireInRange,
-  resolveEffectiveTtsConfig,
-  scheduleCleanup,
-  summarizeText,
-  trimToUndefined,
-  truncateErrorDetail,
-  TTS_AUTO_MODES,
-} from "openclaw/plugin-sdk/speech-core";
-export type {
-  ResolvedTtsConfig,
-  ResolvedTtsModelOverrides,
-  SpeechDirectiveTokenParseContext,
-  SpeechDirectiveTokenParseResult,
-  SpeechListVoicesRequest,
-  SpeechModelOverridePolicy,
-  SpeechProviderConfig,
-  SpeechProviderConfiguredContext,
-  SpeechProviderOverrides,
-  SpeechProviderPlugin,
-  SpeechProviderPreparedSynthesis,
-  SpeechProviderPrepareSynthesisContext,
-  SpeechProviderResolveConfigContext,
-  SpeechProviderResolveTalkConfigContext,
-  SpeechProviderResolveTalkOverridesContext,
-  SpeechSynthesisRequest,
-  SpeechSynthesisStreamRequest,
-  SpeechSynthesisStreamResult,
-  SpeechSynthesisTarget,
-  SpeechTelephonySynthesisRequest,
-  SpeechVoiceOption,
-  TtsConfigResolutionContext,
-  TtsDirectiveOverrides,
-  TtsDirectiveParseResult,
-} from "openclaw/plugin-sdk/speech-core";
diff --git a/packages/speech-core/package.json b/packages/speech-core/package.json
index e27c79f1d190..1fd094355af6 100644
--- a/packages/speech-core/package.json
+++ b/packages/speech-core/package.json
@@ -12,11 +12,6 @@
       "import": "./dist/runtime-api.mjs",
       "default": "./dist/runtime-api.mjs"
     },
-    "./api": {
-      "types": "./dist/api.d.mts",
-      "import": "./dist/api.mjs",
-      "default": "./dist/api.mjs"
-    },
     "./runtime-api": {
       "types": "./dist/runtime-api.d.mts",
       "import": "./dist/runtime-api.mjs",
diff --git a/packages/speech-core/src/tts.test.ts b/packages/speech-core/src/tts.test.ts
index 6f134a02b689..d5be0efad0f7 100644
--- a/packages/speech-core/src/tts.test.ts
+++ b/packages/speech-core/src/tts.test.ts
@@ -83,8 +83,8 @@ vi.mock("openclaw/plugin-sdk/channel-targets", () => ({
   },
 }));
 
-vi.mock("../api.js", async () => {
-  const actual = await vi.importActual<typeof import("../api.js")>("../api.js");
+vi.mock("openclaw/plugin-sdk/speech-core", async () => {
+  const actual = await vi.importActual("openclaw/plugin-sdk/speech-core");
   const mockProvider: SpeechProviderPlugin = {
     id: "mock",
     label: "Mock",
diff --git a/packages/speech-core/src/tts.ts b/packages/speech-core/src/tts.ts
index e330d090ea58..9e72487ae00a 100644
--- a/packages/speech-core/src/tts.ts
+++ b/packages/speech-core/src/tts.ts
@@ -27,17 +27,6 @@ import {
 import { isVerbose, logVerbose } from "openclaw/plugin-sdk/runtime-env";
 import { tempWorkspaceSync, resolvePreferredOpenClawTmpDir } from "openclaw/plugin-sdk/sandbox";
 import { privateFileStoreSync } from "openclaw/plugin-sdk/security-runtime";
-import {
-  normalizeLowercaseStringOrEmpty,
-  normalizeOptionalLowercaseString,
-  normalizeOptionalString,
-} from "openclaw/plugin-sdk/string-coerce-runtime";
-import { stripMarkdown } from "openclaw/plugin-sdk/text-chunking";
-import {
-  resolveConfigDir,
-  resolveUserPath,
-  truncateUtf16Safe,
-} from "openclaw/plugin-sdk/text-utility-runtime";
 import {
   canonicalizeSpeechProviderId,
   getSpeechProvider,
@@ -57,7 +46,18 @@ import {
   type TtsDirectiveOverrides,
   type TtsDirectiveParseResult,
   type TtsConfigResolutionContext,
-} from "../api.js";
+} from "openclaw/plugin-sdk/speech-core";
+import {
+  normalizeLowercaseStringOrEmpty,
+  normalizeOptionalLowercaseString,
+  normalizeOptionalString,
+} from "openclaw/plugin-sdk/string-coerce-runtime";
+import { stripMarkdown } from "openclaw/plugin-sdk/text-chunking";
+import {
+  resolveConfigDir,
+  resolveUserPath,
+  truncateUtf16Safe,
+} from "openclaw/plugin-sdk/text-utility-runtime";
 import { withSpeakerSelectionCompat } from "../speaker.js";
 import {
   resolvePrimaryVoiceProviderCandidate,
diff --git a/tsconfig.json b/tsconfig.json
index ea6ef6b78a4e..6ce69154a170 100644
--- a/tsconfig.json
+++ b/tsconfig.json
@@ -237,7 +237,6 @@
       ],
       "@openclaw/web-content-core/*": ["./packages/web-content-core/src/*"],
       "@openclaw/speech-core": ["./packages/speech-core/runtime-api.ts"],
-      "@openclaw/speech-core/api": ["./packages/speech-core/api.ts"],
       "@openclaw/speech-core/runtime-api": ["./packages/speech-core/runtime-api.ts"],
       "@openclaw/speech-core/speaker": ["./packages/speech-core/speaker.ts"],
       "@openclaw/speech-core/voice-models": ["./packages/speech-core/voice-models.ts"],
diff --git a/tsdown.config.ts b/tsdown.config.ts
index 24e4fccfa3a3..221efd3e2886 100644
--- a/tsdown.config.ts
+++ b/tsdown.config.ts
@@ -394,7 +394,6 @@ function buildPackageDistEntriesFromExports(packageDir: string): Record<string,
 
 function buildSpeechCoreDistEntries(): Record<string, string> {
   return {
-    api: "packages/speech-core/api.ts",
     "runtime-api": "packages/speech-core/runtime-api.ts",
     speaker: "packages/speech-core/speaker.ts",
     "voice-models": "packages/speech-core/voice-models.ts",