@xopcai/xopc 0.0.389 → 0.0.390

package/dist/src/usage/scenario-registry.js

@@ -1,126 +1,5 @@
+import { __esmMin } from "../../_virtual/_rolldown/runtime.js";
 //#region src/usage/scenario-registry.ts
-const SCENARIOS = {
-	"images.generate": {
-		category: "other",
-		reasonKey: "usage.reason.imageGeneration",
-		trigger: "agent"
-	},
-	"audio.transcribe": {
-		category: "other",
-		reasonKey: "usage.reason.audioTranscription",
-		trigger: "user"
-	},
-	"audio.speech": {
-		category: "other",
-		reasonKey: "usage.reason.audioSpeech",
-		trigger: "agent"
-	},
-	"agent.answer": {
-		category: "chat",
-		reasonKey: "usage.reason.agentAnswer",
-		trigger: "user"
-	},
-	"agent.continue_after_tool": {
-		category: "tool_loop",
-		reasonKey: "usage.reason.continueAfterTool",
-		trigger: "agent"
-	},
-	"agent.background_review": {
-		category: "delegation",
-		reasonKey: "usage.reason.backgroundReview",
-		trigger: "agent"
-	},
-	"session.compact": {
-		category: "compaction",
-		reasonKey: "usage.reason.sessionCompaction",
-		trigger: "system"
-	},
-	"session.generate_title": {
-		category: "session_title",
-		reasonKey: "usage.reason.sessionTitle",
-		trigger: "system"
-	},
-	"media.understand_image": {
-		category: "image_understanding",
-		reasonKey: "usage.reason.imageUnderstanding",
-		trigger: "user"
-	},
-	"tool.web_extract": {
-		category: "web_extract",
-		reasonKey: "usage.reason.webExtract",
-		trigger: "agent"
-	},
-	"tool.session_search_summarize": {
-		category: "session_search",
-		reasonKey: "usage.reason.sessionSearch",
-		trigger: "agent"
-	},
-	"note.generate": {
-		category: "note_generation",
-		reasonKey: "usage.reason.noteGeneration",
-		trigger: "user"
-	},
-	"automation.generate_draft": {
-		category: "automation",
-		reasonKey: "usage.reason.automationDraft",
-		trigger: "user"
-	},
-	"scene.execute": {
-		category: "scene",
-		reasonKey: "usage.reason.sceneExecution",
-		trigger: "scheduled"
-	},
-	"task.plan_contract": {
-		category: "task_planning",
-		reasonKey: "usage.reason.taskPlanning",
-		trigger: "agent"
-	},
-	"task.judge_result": {
-		category: "task_judging",
-		reasonKey: "usage.reason.taskJudging",
-		trigger: "agent"
-	},
-	"voice.summarize_for_tts": {
-		category: "voice_summary",
-		reasonKey: "usage.reason.voiceSummary",
-		trigger: "user"
-	},
-	"voice.select": {
-		category: "voice_summary",
-		reasonKey: "usage.reason.voiceSelection",
-		trigger: "user"
-	},
-	"home.generate_advice": {
-		category: "home_intelligence",
-		reasonKey: "usage.reason.homeAdvice",
-		trigger: "system"
-	},
-	"work_discovery.analyze": {
-		category: "work_discovery",
-		reasonKey: "usage.reason.workDiscovery",
-		trigger: "user"
-	},
-	"work_discovery.investigate": {
-		category: "work_discovery",
-		reasonKey: "usage.reason.workInvestigation",
-		trigger: "user"
-	},
-	"discussion.analyze": {
-		category: "discussion_analysis",
-		reasonKey: "usage.reason.discussionAnalysis",
-		trigger: "user"
-	},
-	"text_assist.generate": {
-		category: "other",
-		reasonKey: "usage.reason.textAssist",
-		trigger: "user"
-	},
-	"user_model.interpret": {
-		category: "other",
-		reasonKey: "usage.reason.userModelInterpretation",
-		trigger: "system"
-	}
-};
 function resolveAiUsageScenario(operation) {
 	return SCENARIOS[operation] ?? {
 		category: "other",
@@ -128,5 +7,131 @@ function resolveAiUsageScenario(operation) {
 		trigger: "system"
 	};
 }
+var SCENARIOS;
+var init_scenario_registry = __esmMin((() => {
+	SCENARIOS = {
+		"images.generate": {
+			category: "other",
+			reasonKey: "usage.reason.imageGeneration",
+			trigger: "agent"
+		},
+		"audio.transcribe": {
+			category: "other",
+			reasonKey: "usage.reason.audioTranscription",
+			trigger: "user"
+		},
+		"audio.speech": {
+			category: "other",
+			reasonKey: "usage.reason.audioSpeech",
+			trigger: "agent"
+		},
+		"agent.answer": {
+			category: "chat",
+			reasonKey: "usage.reason.agentAnswer",
+			trigger: "user"
+		},
+		"agent.continue_after_tool": {
+			category: "tool_loop",
+			reasonKey: "usage.reason.continueAfterTool",
+			trigger: "agent"
+		},
+		"agent.background_review": {
+			category: "delegation",
+			reasonKey: "usage.reason.backgroundReview",
+			trigger: "agent"
+		},
+		"session.compact": {
+			category: "compaction",
+			reasonKey: "usage.reason.sessionCompaction",
+			trigger: "system"
+		},
+		"session.generate_title": {
+			category: "session_title",
+			reasonKey: "usage.reason.sessionTitle",
+			trigger: "system"
+		},
+		"media.understand_image": {
+			category: "image_understanding",
+			reasonKey: "usage.reason.imageUnderstanding",
+			trigger: "user"
+		},
+		"tool.web_extract": {
+			category: "web_extract",
+			reasonKey: "usage.reason.webExtract",
+			trigger: "agent"
+		},
+		"tool.session_search_summarize": {
+			category: "session_search",
+			reasonKey: "usage.reason.sessionSearch",
+			trigger: "agent"
+		},
+		"note.generate": {
+			category: "note_generation",
+			reasonKey: "usage.reason.noteGeneration",
+			trigger: "user"
+		},
+		"automation.generate_draft": {
+			category: "automation",
+			reasonKey: "usage.reason.automationDraft",
+			trigger: "user"
+		},
+		"scene.execute": {
+			category: "scene",
+			reasonKey: "usage.reason.sceneExecution",
+			trigger: "scheduled"
+		},
+		"task.plan_contract": {
+			category: "task_planning",
+			reasonKey: "usage.reason.taskPlanning",
+			trigger: "agent"
+		},
+		"task.judge_result": {
+			category: "task_judging",
+			reasonKey: "usage.reason.taskJudging",
+			trigger: "agent"
+		},
+		"voice.summarize_for_tts": {
+			category: "voice_summary",
+			reasonKey: "usage.reason.voiceSummary",
+			trigger: "user"
+		},
+		"voice.select": {
+			category: "voice_summary",
+			reasonKey: "usage.reason.voiceSelection",
+			trigger: "user"
+		},
+		"home.generate_advice": {
+			category: "home_intelligence",
+			reasonKey: "usage.reason.homeAdvice",
+			trigger: "system"
+		},
+		"work_discovery.analyze": {
+			category: "work_discovery",
+			reasonKey: "usage.reason.workDiscovery",
+			trigger: "user"
+		},
+		"work_discovery.investigate": {
+			category: "work_discovery",
+			reasonKey: "usage.reason.workInvestigation",
+			trigger: "user"
+		},
+		"discussion.analyze": {
+			category: "discussion_analysis",
+			reasonKey: "usage.reason.discussionAnalysis",
+			trigger: "user"
+		},
+		"text_assist.generate": {
+			category: "other",
+			reasonKey: "usage.reason.textAssist",
+			trigger: "user"
+		},
+		"user_model.interpret": {
+			category: "other",
+			reasonKey: "usage.reason.userModelInterpretation",
+			trigger: "system"
+		}
+	};
+}));
 //#endregion
-export { resolveAiUsageScenario };
+init_scenario_registry();
+export { init_scenario_registry, resolveAiUsageScenario };

package/dist/src/voice/dashscope/streaming-stt-session.js

@@ -1,6 +1,8 @@
+import { assertProviderUsage, init_model_usage } from "../../providers/model-usage.js";
 import { createRequire } from "node:module";
 import crypto from "node:crypto";
 //#region src/voice/dashscope/streaming-stt-session.ts
+init_model_usage();
 const { WebSocket: NodeWebSocket } = createRequire(import.meta.url)("ws");
 const DEFAULT_REALTIME_URL = "wss://dashscope.aliyuncs.com/api-ws/v1/inference";
 const MAX_BUFFERED_AUDIO_BYTES = 512 * 1024;
@@ -32,6 +34,7 @@ async function openDashScopeStreamingStt(request) {
 	if (request.signal.aborted) throw request.signal.reason instanceof Error ? request.signal.reason : /* @__PURE__ */ new Error("DashScope realtime STT aborted");
 	if (request.inputFormat.encoding !== "pcm_s16le" || request.inputFormat.channels !== 1) throw new Error("DashScope realtime STT requires mono PCM16 audio");
 	const taskId = crypto.randomUUID();
+	assertProviderUsage("alibaba", request.baseUrl);
 	const socket = new NodeWebSocket(realtimeUrl(request.baseUrl), {
 		headers: {
 			Authorization: `Bearer ${request.apiKey}`,

package/dist/src/voice/dashscope/streaming-tts-stream.js

@@ -1,9 +1,11 @@
 import { createLogger } from "../../utils/logger/index.js";
 import { init_logger } from "../../utils/logger.js";
+import { assertProviderUsage, init_model_usage } from "../../providers/model-usage.js";
 import { createRequire } from "node:module";
 import crypto from "node:crypto";
 import { setTimeout as setTimeout$1 } from "node:timers/promises";
 //#region src/voice/dashscope/streaming-tts-stream.ts
+init_model_usage();
 init_logger();
 const { WebSocket } = createRequire(import.meta.url)("ws");
 const DEFAULT_REALTIME_URL = "wss://dashscope.aliyuncs.com/api-ws/v1/realtime";
@@ -93,6 +95,7 @@ async function openStreamingTtsAttempt(request) {
 	if (!request.apiKey) throw new Error("DashScope TTS API key is unavailable");
 	request.signal.throwIfAborted();
 	const startedAt = Date.now();
+	assertProviderUsage("alibaba", request.baseUrl);
 	const socket = new WebSocket(realtimeUrl(request.baseUrl, request.model), {
 		headers: { Authorization: `Bearer ${request.apiKey}` },
 		perMessageDeflate: false

package/dist/src/voice/dictation.d.ts

@@ -0,0 +1,12 @@
+import type { Config } from '../config/schema.js';
+/** Select before capture; a failed request must never switch billing services. */
+export declare function resolveDictationRoute(config: Config): {
+    mode: "unavailable";
+    provider?: undefined;
+} | {
+    mode: "streaming";
+    provider: string;
+} | {
+    mode: "recorded";
+    provider: string;
+};

package/dist/src/voice/dictation.js

@@ -0,0 +1,22 @@
+import { resolveSTTProviderChain } from "./stt/factory.js";
+import { mergeSttConfigFromAppConfig } from "./stt/config.js";
+import "./stt/index.js";
+import { resolveStreamingStt } from "./realtime/runtime.js";
+//#region src/voice/dictation.ts
+/** Select before capture; a failed request must never switch billing services. */
+function resolveDictationRoute(config) {
+	const primary = resolveSTTProviderChain(mergeSttConfigFromAppConfig(config.tools?.media?.audio, config.tools?.media))[0];
+	const streaming = config.voice?.realtime?.enabled ? resolveStreamingStt(config) : void 0;
+	if (config.voice?.realtime?.enabled && config.voice.realtime.stt && !streaming) return { mode: "unavailable" };
+	if (streaming && (config.voice?.realtime?.stt || streaming.route.provider === primary?.id)) return {
+		mode: "streaming",
+		provider: streaming.route.provider
+	};
+	if (primary) return {
+		mode: "recorded",
+		provider: primary.id
+	};
+	return { mode: "unavailable" };
+}
+//#endregion
+export { resolveDictationRoute };

package/dist/src/voice/metadata/builtin.js

@@ -457,6 +457,60 @@ const builtinVoiceProviderMetadata = [
 		}
 	}
 ];
+builtinVoiceProviderMetadata.push({
+	id: "qwen-token-plan-cn",
+	capability: "tts",
+	displayName: "Alibaba Bailian Subscription",
+	description: "Subscription speech synthesis using the saved model-service credential.",
+	models: [{
+		id: "qwen-audio-3.0-tts-plus",
+		name: "Qwen Audio 3.0 TTS Plus"
+	}],
+	voices: [{
+		id: "longanhuan_v3.6",
+		name: "Long Anhuan"
+	}],
+	fields: [{
+		key: "model",
+		label: "Model",
+		type: "select",
+		options: [{
+			id: "qwen-audio-3.0-tts-plus",
+			name: "Qwen Audio 3.0 TTS Plus"
+		}]
+	}, {
+		key: "voice",
+		label: "Voice",
+		type: "string",
+		defaultValue: "longanhuan_v3.6"
+	}],
+	diagnostics: {
+		requiresApiKey: false,
+		configPath: "messages.tts.providers.qwen-token-plan-cn"
+	}
+}, {
+	id: "qwen-token-plan-cn",
+	capability: "stt",
+	displayName: "Alibaba Bailian Subscription",
+	description: "Recorded-audio recognition using the saved model-service credential. Streaming dictation is not configured by this adapter.",
+	models: [{
+		id: "qwen-audio-3.0-asr-flash",
+		name: "Qwen Audio 3.0 ASR Flash"
+	}],
+	fields: [{
+		key: "model",
+		label: "Model",
+		type: "select",
+		options: [{
+			id: "qwen-audio-3.0-asr-flash",
+			name: "Qwen Audio 3.0 ASR Flash"
+		}]
+	}],
+	diagnostics: {
+		requiresApiKey: false,
+		configPath: "tools.media.audio.providers.qwen-token-plan-cn"
+	}
+});
 for (const metadata of builtinVoiceProviderMetadata) registerVoiceProviderMetadata(metadata);
 //#endregion
 export { builtinVoiceProviderMetadata };

package/dist/src/voice/platform-streams.js

@@ -1,5 +1,7 @@
+import { assertProviderUsage, init_model_usage } from "../providers/model-usage.js";
 import WebSocket from "ws";
 //#region src/voice/platform-streams.ts
+init_model_usage();
 const MAX_BUFFERED_BYTES = 512 * 1024;
 const MAX_ERROR_BODY_BYTES = 16 * 1024;
 var PlatformVoiceError = class extends Error {
@@ -22,6 +24,7 @@ var PlatformVoiceError = class extends Error {
 };
 function connect(url, token, signal, onEvent, onFailure) {
 	signal.throwIfAborted();
+	assertProviderUsage("xopc-cloud", String(url));
 	const socket = new WebSocket(url, {
 		headers: { Authorization: `Bearer ${token}` },
 		perMessageDeflate: false,

package/dist/src/voice/realtime/omniEngine.js

@@ -1,5 +1,6 @@
 import { createLogger } from "../../utils/logger/index.js";
 import { init_logger } from "../../utils/logger.js";
+import { assertProviderUsage, init_model_usage } from "../../providers/model-usage.js";
 import { AudioPlaybackWindow } from "./audio-playback-window.js";
 import { TurnCoordinator } from "./turnPolicy.js";
 import { PlaybackEchoCandidates } from "./playback-echo.js";
@@ -7,6 +8,7 @@ import { PcmFrameBuffer } from "./pcmFrameBuffer.js";
 import crypto from "node:crypto";
 import WebSocket from "ws";
 //#region src/voice/realtime/omniEngine.ts
+init_model_usage();
 init_logger();
 const log = createLogger("Voice:Omni");
 const MAX_QUEUED_AUDIO_BYTES = 24e3 * 2 * 60;
@@ -289,6 +291,7 @@ function createOmniVoiceEngine(options) {
 		start() {
 			return new Promise((resolve, reject) => {
 				rejectStart = reject;
+				assertProviderUsage(options.route.route.provider, options.route.url);
 				socket = new WebSocket(options.route.url, {
 					headers: { Authorization: `Bearer ${options.route.apiKey}` },
 					maxPayload: 1024 * 1024,

package/dist/src/voice/realtime/omniRoute.js

@@ -3,6 +3,7 @@ import { init_voice } from "../../../packages/realtime-protocol/src/voice.js";
 import { init_xopc_cloud_config, resolveXopcModelRouterUrl } from "../../providers/xopc-cloud-config.js";
 import { getProviderAuthService, init_provider_auth_service } from "../../providers/provider-auth-service.js";
 import { requirePlatformVoiceModel } from "../platform-catalog.js";
+import { BAILIAN_PLAN_HOST } from "../token-plan.js";
 //#region src/voice/realtime/omniRoute.ts
 init_voice();
 init_provider_auth_service();
@@ -19,11 +20,11 @@ async function resolveOmniRoute(config) {
 		if (!model.voice.voices.some((voice) => voice.id === slice.voice && voiceSupportsMode(voice, "conversation"))) throw new Error("Selected voice is unavailable");
 	}
 	const managed = slice.provider === "xopc-cloud";
-	const url = new URL(managed ? `${resolveXopcModelRouterUrl(slice.baseUrl)}/audio/conversations/realtime` : slice.baseUrl ?? "wss://dashscope.aliyuncs.com/api-ws/v1/realtime");
+	const url = new URL(managed ? `${resolveXopcModelRouterUrl(slice.baseUrl)}/audio/conversations/realtime` : slice.provider === "qwen-token-plan-cn" ? `${BAILIAN_PLAN_HOST.replace("https:", "wss:")}/api-ws/v1/realtime` : slice.baseUrl ?? "wss://dashscope.aliyuncs.com/api-ws/v1/realtime");
 	if (managed && url.protocol === "https:") url.protocol = "wss:";
 	if (url.protocol !== "wss:" || url.username || url.password || url.hash || url.port) throw new Error("Omni requires a secure WebSocket endpoint");
 	if (!managed && (!(new Set(["dashscope.aliyuncs.com", "dashscope-intl.aliyuncs.com"]).has(url.hostname) || /^[a-zA-Z0-9-]+\.(cn-beijing|ap-southeast-1)\.maas\.aliyuncs\.com$/.test(url.hostname)) || url.pathname !== "/api-ws/v1/realtime")) throw new Error("Unsupported DashScope realtime endpoint");
-	const apiKey = managed ? await getProviderAuthService().resolveApiKey("xopc-cloud") : slice.apiKey ?? config.tools?.media?.audio?.providers?.alibaba?.apiKey ?? await getProviderAuthService().resolveApiKey("dashscope");
+	const apiKey = managed || slice.provider === "qwen-token-plan-cn" ? await getProviderAuthService().resolveApiKey(slice.provider) : slice.apiKey ?? config.tools?.media?.audio?.providers?.alibaba?.apiKey ?? await getProviderAuthService().resolveApiKey("dashscope");
 	if (!apiKey) throw new Error("Omni credentials are unavailable");
 	url.search = new URLSearchParams({ model: slice.model }).toString();
 	if (serviceVersion) url.searchParams.set("service_version", String(serviceVersion));

package/dist/src/voice/realtime/runtime.js

@@ -1,6 +1,7 @@
 import { createLogger } from "../../utils/logger/index.js";
 import { init_logger } from "../../utils/logger.js";
 import { init_changes, onUserContextChange } from "../../user-context/changes.js";
+import { init_model_usage, withModelUsageOrigin } from "../../providers/model-usage.js";
 import { decodeVoiceUplinkAudioFrame, encodeVoiceAudioFrame } from "../../../packages/realtime-protocol/src/voice-audio.js";
 import { VOICE_REALTIME_HEARTBEAT_INTERVAL_MS, VOICE_REALTIME_MAX_BINARY_FRAME_BYTES, VOICE_REALTIME_START_TIMEOUT_MS, VOICE_REALTIME_WS_PATH, init_voice, parseVoiceClientJsonFrame, parseVoiceClientMessage } from "../../../packages/realtime-protocol/src/voice.js";
 import { getModelCatalogStore, init_model_catalog_store } from "../../providers/model-catalog-store.js";
@@ -22,6 +23,7 @@ import { applyVoiceOutputGain } from "./pcm-output-gain.js";
 import { createRequire } from "node:module";
 import crypto from "node:crypto";
 //#region src/voice/realtime/runtime.ts
+init_model_usage();
 init_voice();
 init_logger();
 init_changes();
@@ -655,7 +657,7 @@ var VoiceRealtimeRuntime = class {
 			}
 		};
 		const clientMetrics = /* @__PURE__ */ new Set();
-		socket.on("message", (data, isBinary) => {
+		socket.on("message", (data, isBinary) => withModelUsageOrigin("interactive", () => {
 			lastActivityAt = Date.now();
 			if (isBinary) {
 				if (!claim) {
@@ -879,7 +881,7 @@ var VoiceRealtimeRuntime = class {
 				message: "Voice session is already started",
 				recoverable: true
 			});
-		});
+		}));
 		socket.once("close", () => void shutdown("disconnected", false));
 		socket.once("error", (err) => log.debug({ err }, "Voice client socket failed"));
 	}

package/dist/src/voice/stt/factory.js

@@ -1,5 +1,6 @@
 import { createLogger } from "../../utils/logger/index.js";
 import { init_logger } from "../../utils/logger.js";
+import { init_profiles, isInteractiveOnlyProvider } from "../../providers/setup/profiles.js";
 import { getModelCatalogStore, init_model_catalog_store } from "../../providers/model-catalog-store.js";
 import { compareCatalogModels } from "../../providers/model-catalog-ranking.js";
 import { getMediaUnderstandingProvider, listMediaUnderstandingProviders } from "../../media-understanding/registry.js";
@@ -9,6 +10,7 @@ import { readSttProviderFields, resolveSttProviderConfigSlice } from "./config-s
 //#region src/voice/stt/factory.ts
 init_logger();
 init_model_catalog_store();
+init_profiles();
 const log = createLogger("STT:Factory");
 function resolveApiKey(providerId, fields, envKey) {
 	return fields.apiKey ?? (envKey ? process.env[envKey] : void 0);
@@ -58,16 +60,20 @@ function resolveSTTProviderOrder(primary, fallback, config) {
 		return order;
 	}
 	if (!config.managedAuto) return [primary];
-	const configured = sortAudioProvidersByAutoPriority(listMediaUnderstandingProviders()).filter((providerId) => resolveSTTProviderConfig(providerId, config) !== null);
+	const configured = sortAudioProvidersByAutoPriority(listMediaUnderstandingProviders()).filter((providerId) => {
+		const resolved = resolveSTTProviderConfig(providerId, config);
+		return resolved !== null && !isInteractiveOnlyProvider(providerId, resolved.baseUrl);
+	});
 	const order = [primary];
 	for (const providerId of configured) if (providerId !== primary && !order.includes(providerId)) order.push(providerId);
 	return order;
 }
 function resolveModelEntryChain(config) {
+	const pinned = Boolean(config.models?.length) && config.fallback?.enabled === false && !config.managedAuto;
 	const entries = resolveModelEntries({
 		capability: "audio",
 		capabilityModels: config.models,
-		sharedModels: config.sharedModels
+		sharedModels: pinned ? void 0 : config.sharedModels
 	});
 	const chain = [];
 	for (const entry of entries) {
@@ -76,7 +82,7 @@ function resolveModelEntryChain(config) {
 		const resolved = resolveSTTProviderConfig(providerId, config, entry);
 		if (resolved) chain.push(resolved);
 	}
-	if (chain.some((entry) => entry.id === "xopc-cloud")) appendManagedCloudModels(chain, config);
+	if (!pinned && chain.some((entry) => entry.id === "xopc-cloud")) appendManagedCloudModels(chain, config);
 	return chain;
 }
 function appendManagedCloudModels(chain, config) {

package/dist/src/voice/stt/providers/bailian-plan-transcription.d.ts

@@ -0,0 +1,2 @@
+import type { MediaUnderstandingProvider } from '../../../media-understanding/types.js';
+export declare const bailianPlanTranscriptionProvider: MediaUnderstandingProvider;

package/dist/src/voice/stt/providers/bailian-plan-transcription.js

@@ -0,0 +1,32 @@
+import { assertProviderUsage, init_model_usage } from "../../../providers/model-usage.js";
+import { init_sync_provider_auth, resolveProviderApiKeyForAgentSync } from "../../../auth/sync-provider-auth.js";
+import { getProviderAuthService, init_provider_auth_service } from "../../../providers/provider-auth-service.js";
+import { registerMediaUnderstandingProvider } from "../../../media-understanding/registry.js";
+import { alibabaTranscriptionProvider } from "./alibaba-transcription.js";
+import { BAILIAN_PLAN_HOST, BAILIAN_PLAN_PROVIDER, BAILIAN_PLAN_STT_MODEL } from "../../token-plan.js";
+//#region src/voice/stt/providers/bailian-plan-transcription.ts
+init_sync_provider_auth();
+init_provider_auth_service();
+init_model_usage();
+const bailianPlanTranscriptionProvider = {
+	id: BAILIAN_PLAN_PROVIDER,
+	capabilities: ["audio"],
+	requiresApiKey: false,
+	defaultModels: { audio: BAILIAN_PLAN_STT_MODEL },
+	isConfigured: () => Boolean(resolveProviderApiKeyForAgentSync("qwen-token-plan-cn") ?? process.env.QWEN_TOKEN_PLAN_CN_API_KEY),
+	transcribeAudio: async (request) => {
+		assertProviderUsage(BAILIAN_PLAN_PROVIDER);
+		const apiKey = await getProviderAuthService().resolveApiKey(BAILIAN_PLAN_PROVIDER);
+		if (!apiKey) throw new Error("Bailian subscription STT credential is unavailable");
+		if (request.model && request.model !== "qwen-audio-3.0-asr-flash") throw new Error("Unsupported Bailian subscription STT model");
+		return alibabaTranscriptionProvider.transcribeAudio({
+			...request,
+			apiKey,
+			model: BAILIAN_PLAN_STT_MODEL,
+			baseUrl: `${BAILIAN_PLAN_HOST}/api/v1/services/aigc/multimodal-generation/generation`
+		});
+	}
+};
+registerMediaUnderstandingProvider(bailianPlanTranscriptionProvider);
+//#endregion
+export { bailianPlanTranscriptionProvider };

package/dist/src/voice/stt/providers/index.d.ts

@@ -13,3 +13,4 @@
 export { openAiTranscriptionProvider } from './openai-transcription.js';
 export { alibabaTranscriptionProvider } from './alibaba-transcription.js';
 export { xopcCloudTranscriptionProvider } from './xopc-cloud-transcription.js';
+export { bailianPlanTranscriptionProvider } from './bailian-plan-transcription.js';

package/dist/src/voice/stt/providers/index.js

@@ -1,4 +1,5 @@
 import { openAiTranscriptionProvider } from "./openai-transcription.js";
 import { alibabaTranscriptionProvider } from "./alibaba-transcription.js";
 import { xopcCloudTranscriptionProvider } from "./xopc-cloud-transcription.js";
-export { alibabaTranscriptionProvider, openAiTranscriptionProvider, xopcCloudTranscriptionProvider };
+import { bailianPlanTranscriptionProvider } from "./bailian-plan-transcription.js";
+export { alibabaTranscriptionProvider, bailianPlanTranscriptionProvider, openAiTranscriptionProvider, xopcCloudTranscriptionProvider };

package/dist/src/voice/token-plan.d.ts

@@ -0,0 +1,8 @@
+/** Verified Beijing individual-plan bindings; ordinary DashScope settings stay separate. */
+export declare const BAILIAN_PLAN_PROVIDER = "qwen-token-plan-cn";
+export declare const BAILIAN_PLAN_HOST = "https://token-plan.cn-beijing.maas.aliyuncs.com";
+export declare const BAILIAN_PLAN_TTS_MODEL = "qwen-audio-3.0-tts-plus";
+export declare const BAILIAN_PLAN_STT_MODEL = "qwen-audio-3.0-asr-flash";
+export declare const BAILIAN_PLAN_OMNI_MODEL = "qwen-audio-3.0-realtime-plus";
+export declare const BAILIAN_PLAN_TTS_VOICE = "longanhuan_v3.6";
+export declare const BAILIAN_PLAN_OMNI_VOICE = "longanqian";

package/dist/src/voice/token-plan.js

@@ -0,0 +1,11 @@
+//#region src/voice/token-plan.ts
+/** Verified Beijing individual-plan bindings; ordinary DashScope settings stay separate. */
+const BAILIAN_PLAN_PROVIDER = "qwen-token-plan-cn";
+const BAILIAN_PLAN_HOST = "https://token-plan.cn-beijing.maas.aliyuncs.com";
+const BAILIAN_PLAN_TTS_MODEL = "qwen-audio-3.0-tts-plus";
+const BAILIAN_PLAN_STT_MODEL = "qwen-audio-3.0-asr-flash";
+const BAILIAN_PLAN_OMNI_MODEL = "qwen-audio-3.0-realtime-plus";
+const BAILIAN_PLAN_TTS_VOICE = "longanhuan_v3.6";
+const BAILIAN_PLAN_OMNI_VOICE = "longanqian";
+//#endregion
+export { BAILIAN_PLAN_HOST, BAILIAN_PLAN_OMNI_MODEL, BAILIAN_PLAN_OMNI_VOICE, BAILIAN_PLAN_PROVIDER, BAILIAN_PLAN_STT_MODEL, BAILIAN_PLAN_TTS_MODEL, BAILIAN_PLAN_TTS_VOICE };

package/dist/src/voice/tts/factory.js

@@ -1,5 +1,6 @@
 import { createLogger } from "../../utils/logger/index.js";
 import { init_logger } from "../../utils/logger.js";
+import { init_profiles, isInteractiveOnlyProvider } from "../../providers/setup/profiles.js";
 import { getModelCatalogStore, init_model_catalog_store } from "../../providers/model-catalog-store.js";
 import { compareCatalogModels } from "../../providers/model-catalog-ranking.js";
 import { getSpeechProvider, listSpeechProviders } from "./speech-registry.js";
@@ -8,6 +9,7 @@ import { buildTtsResolveRawConfig } from "./config-slice.js";
 //#region src/voice/tts/factory.ts
 init_logger();
 init_model_catalog_store();
+init_profiles();
 const log = createLogger("TTS:Factory");
 /** Resolve a single provider id → plugin + per-call config, or null if unavailable. */
 function resolveSpeechProvider(providerId, config, providerOverride) {
@@ -50,7 +52,8 @@ function resolveProviderOrder(primary, fallback, config) {
 	}
 	if (!config.managedAuto) return [primary];
 	const configured = listSpeechProviders().map((plugin) => {
-		return resolveSpeechProvider(plugin.id, config) ? plugin : null;
+		const resolved = resolveSpeechProvider(plugin.id, config);
+		return resolved && !isInteractiveOnlyProvider(plugin.id, resolved.providerConfig.baseUrl) ? plugin : null;
 	}).filter((plugin) => plugin !== null).sort((left, right) => {
 		const leftOrder = left.autoSelectOrder ?? Number.MAX_SAFE_INTEGER;
 		const rightOrder = right.autoSelectOrder ?? Number.MAX_SAFE_INTEGER;

package/dist/src/voice/tts/providers/bailian-plan-speech.d.ts

@@ -0,0 +1,2 @@
+import type { SpeechProviderPlugin } from '../speech-provider-types.js';
+export declare const bailianPlanSpeechProvider: SpeechProviderPlugin;