@mastra/voice-speechify 0.13.0 → 0.14.0

package/CHANGELOG.md

@@ -1,5 +1,91 @@
 # @mastra/voice-speechify
 
+## 0.14.0
+
+### Minor Changes
+
+- Added support for Speechify's Simba 3.2 and Simba 3.0 text-to-speech models. Simba 3.2 is Speechify's latest streaming model with lower latency and richer expressivity, and is the recommended model for English speech. ([#19411](https://github.com/mastra-ai/mastra/pull/19411))
+
+  ```typescript
+  import { SpeechifyVoice } from '@mastra/voice-speechify';
+
+  // Set as the default model
+  const voice = new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+  });
+
+  // Or override per request
+  const stream = await voice.speak('Hello world', { model: 'simba-3.2' });
+  ```
+
+  Note: `simba-3.2` and `simba-3.0` are currently English only. The default model remains `simba-english`.
+
+### Patch Changes
+
+- Fixed voice selection for the Simba 3 models. Speechify's `simba-3.2` and `simba-3.0` serve a curated voice set only (`beatrice_32`, `dominic_32`, `edmund_32`, `geffen_32`, `harper_32`, `hugh_32`, `imogen_32`, `wyatt_32`), so pairing them with a classic catalog voice like `george` failed with an API error. ([#19415](https://github.com/mastra-ai/mastra/pull/19415))
+  - Added the curated Simba 3 voices to the voice list, so they type-check as `speaker` and appear in `getSpeakers()`
+  - The default speaker now follows the configured model: `harper_32` for Simba 3 models, `george` otherwise
+
+  ```typescript
+  import { SpeechifyVoice } from '@mastra/voice-speechify';
+
+  // Works out of the box now — defaults to the harper_32 voice
+  const voice = new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+  });
+
+  // Or pick a curated voice explicitly
+  new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+    speaker: 'imogen_32',
+  });
+  ```
+
+  When overriding the model per request, pass a matching speaker too: `voice.speak('Hi', { model: 'simba-3.2', speaker: 'harper_32' })`.
+
+## 0.14.0-alpha.0
+
+### Minor Changes
+
+- Added support for Speechify's Simba 3.2 and Simba 3.0 text-to-speech models. Simba 3.2 is Speechify's latest streaming model with lower latency and richer expressivity, and is the recommended model for English speech. ([#19411](https://github.com/mastra-ai/mastra/pull/19411))
+
+  ```typescript
+  import { SpeechifyVoice } from '@mastra/voice-speechify';
+
+  // Set as the default model
+  const voice = new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+  });
+
+  // Or override per request
+  const stream = await voice.speak('Hello world', { model: 'simba-3.2' });
+  ```
+
+  Note: `simba-3.2` and `simba-3.0` are currently English only. The default model remains `simba-english`.
+
+### Patch Changes
+
+- Fixed voice selection for the Simba 3 models. Speechify's `simba-3.2` and `simba-3.0` serve a curated voice set only (`beatrice_32`, `dominic_32`, `edmund_32`, `geffen_32`, `harper_32`, `hugh_32`, `imogen_32`, `wyatt_32`), so pairing them with a classic catalog voice like `george` failed with an API error. ([#19415](https://github.com/mastra-ai/mastra/pull/19415))
+  - Added the curated Simba 3 voices to the voice list, so they type-check as `speaker` and appear in `getSpeakers()`
+  - The default speaker now follows the configured model: `harper_32` for Simba 3 models, `george` otherwise
+
+  ```typescript
+  import { SpeechifyVoice } from '@mastra/voice-speechify';
+
+  // Works out of the box now — defaults to the harper_32 voice
+  const voice = new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+  });
+
+  // Or pick a curated voice explicitly
+  new SpeechifyVoice({
+    speechModel: { name: 'simba-3.2' },
+    speaker: 'imogen_32',
+  });
+  ```
+
+  When overriding the model per request, pass a matching speaker too: `voice.speak('Hi', { model: 'simba-3.2', speaker: 'harper_32' })`.
+
 ## 0.13.0
 
 ### Minor Changes

package/README.md

@@ -23,10 +23,10 @@ import { SpeechifyVoice } from '@mastra/voice-speechify';
 
 const voice = new SpeechifyVoice({
   speechModel: {
-    name: 'simba-english', // Optional, defaults to 'simba-english'
+    name: 'simba-3.2', // Optional, defaults to 'simba-english'
     apiKey: 'your-api-key', // Optional, can use SPEECHIFY_API_KEY env var
   },
-  speaker: 'george', // Optional, defaults to 'george'
+  speaker: 'harper_32', // Optional, defaults to a voice that matches the model
 });
 
 // List available speakers
@@ -34,7 +34,7 @@ const speakers = await voice.getSpeakers();
 
 // Generate speech
 const stream = await voice.speak('Hello world', {
-  speaker: 'george', // Optional, defaults to constructor speaker
+  speaker: 'harper_32', // Optional, defaults to constructor speaker
   // Additional Speechify options
   audioFormat: 'mp3',
 });
@@ -49,7 +49,7 @@ The `SpeechifyVoice` constructor accepts the following options:
 
 ```typescript
 interface SpeechifyConfig {
-  name?: string; // Optional Speechify model name (default: 'simba-english')
+  name?: SpeechifyModel; // Optional Speechify model name (default: 'simba-english')
   apiKey?: string; // Optional API key (can also use env var)
 }
 
@@ -59,6 +59,19 @@ new SpeechifyVoice({
 })
 ```
 
+## Available Models
+
+- `simba-3.2`: Speechify's latest streaming-native model with the lowest latency and richest expressivity. Recommended for English. Currently English only.
+- `simba-3.0`: The earlier Simba 3 model, still available. Currently English only.
+- `simba-english`: The default model, optimized for English.
+- `simba-multilingual`: Optimized for non-English or mixed-language input.
+
+The model can also be overridden per request. When overriding to a Simba 3 model, pass a matching speaker too:
+
+```typescript
+const stream = await voice.speak('Hello world', { model: 'simba-3.2', speaker: 'harper_32' });
+```
+
 ## Available Speakers
 
 You can get a list of available speakers:
@@ -66,3 +79,10 @@ You can get a list of available speakers:
 ```typescript
 const speakers = await voice.getSpeakers();
 ```
+
+Voice availability depends on the model:
+
+- `simba-3.2` and `simba-3.0` serve a curated voice set only: `beatrice_32`, `dominic_32`, `edmund_32`, `geffen_32`, `harper_32`, `hugh_32`, `imogen_32`, `wyatt_32`. `simba-3.2` also accepts cloned voices approved by Speechify.
+- `simba-english` and `simba-multilingual` serve the full classic catalog (`george`, `henry`, `carly`, ...) and self-serve cloned voices.
+
+The default speaker follows the configured model: `harper_32` for the Simba 3 models, otherwise `george`.

package/dist/_types/@internal_voice/dist/_types/@internal_ai-sdk-v5/dist/index.d.ts

@@ -231,6 +231,7 @@ declare namespace _ai_sdk_provider_utils {
         Validator,
         asSchema,
         asValidator,
+        cancelResponseBody,
         combineHeaders,
         convertAsyncIteratorToReadableStream,
         convertBase64ToUint8Array,
@@ -249,13 +250,16 @@ declare namespace _ai_sdk_provider_utils {
         dynamicTool,
         executeTool,
         extractResponseHeaders,
+        fetchWithValidatedRedirects,
         generateId,
         getErrorMessage,
         getFromApi,
         getRuntimeEnvironmentUserAgent,
         injectJsonInstructionIntoMessages,
         isAbortError,
+        isBrowserRuntime,
         isParsableJson,
+        isSameOrigin,
         isUrlSupported,
         isValidator,
         jsonSchema,
@@ -494,6 +498,20 @@ export declare type CallSettings = {
  */
 export declare type CallWarning = LanguageModelV2CallWarning;
 
+/**
+ * Cancels a response body to release the underlying connection.
+ *
+ * When a fetch Response is rejected without consuming its body (e.g. a failed
+ * status code, an open-redirect rejection, or a Content-Length that exceeds the
+ * size limit), the underlying TCP socket is not returned to the connection pool
+ * and may stay open until the process runs out of file descriptors. Cancelling
+ * the body avoids this leak.
+ *
+ * Errors thrown while cancelling are ignored: the body may already be locked,
+ * disturbed, or absent, none of which should mask the original rejection.
+ */
+declare function cancelResponseBody(response: Response): Promise<void>;
+
 export declare interface ChatInit<UI_MESSAGE extends UIMessage> {
     /**
      * A unique identifier for the chat. If not provided, a random one will be
@@ -956,7 +974,21 @@ export declare type CreateUIMessage<UI_MESSAGE extends UIMessage> = Omit<UI_MESS
     role?: UI_MESSAGE['role'];
 };
 
-export declare function createUIMessageStream<UI_MESSAGE extends UIMessage>({ execute, onError, originalMessages, onFinish, generateId, }: {
+/**
+ * Creates a UI message stream that can be used to send messages to the client.
+ *
+ * @param options.execute - A function that is called with a writer to write UI message chunks to the stream.
+ * @param options.onError - A function that extracts an error message from an error. Defaults to `() => 'An error occurred.'` so server-side error details are not leaked to the client; supply your own to surface richer messages.
+ * @param options.originalMessages - The original messages. If provided, persistence mode is assumed
+ *   and a message ID is provided for the response message.
+ * @param options.onStepFinish - A callback that is called when each step finishes. Useful for persisting intermediate messages.
+ * @param options.onFinish - A callback that is called when the stream finishes.
+ * @param options.generateId - A function that generates a unique ID. Defaults to the built-in ID generator.
+ *
+ * @returns A `ReadableStream` of UI message chunks.
+ */
+export declare function createUIMessageStream<UI_MESSAGE extends UIMessage>({ execute, onError, // prevent leaking server error details to the client by default
+    originalMessages, onFinish, generateId, }: {
     execute: (options: {
         writer: UIMessageStreamWriter<UI_MESSAGE>;
     }) => Promise<void> | void;
@@ -2129,6 +2161,36 @@ declare function extractResponseHeaders(response: Response): {
  */
 declare type FetchFunction = typeof globalThis.fetch;
 
+/**
+ * Fetches a URL while enforcing the SSRF download guard on every hop.
+ *
+ * Redirects are followed manually (`redirect: 'manual'`) so each hop is
+ * validated with {@link validateDownloadUrl} *before* it is requested. Relying
+ * on the default `redirect: 'follow'` would issue the request to a redirect
+ * target (e.g. an internal address) before we ever see its URL, defeating the
+ * SSRF guard.
+ *
+ * A `redirect: 'manual'` request yields an unreadable opaque response in the
+ * browser (and in other spec-compliant fetch implementations), so the redirect
+ * target cannot be validated here. In a real browser this is safe to follow
+ * natively because SSRF is not reachable (fetch is constrained by CORS and
+ * cannot reach a server's internal network or cloud-metadata). On any other
+ * runtime we cannot validate the hop, so we fail closed rather than follow it
+ * blindly and bypass the SSRF guard.
+ *
+ * The returned response is the final (non-redirect) response. The caller is
+ * responsible for checking `response.ok` and reading the body.
+ *
+ * @throws DownloadError if a hop is unsafe, the redirect limit is exceeded, or
+ * a redirect cannot be validated on a non-browser runtime.
+ */
+declare function fetchWithValidatedRedirects({ url, headers, abortSignal, maxRedirects, }: {
+    url: string;
+    headers?: HeadersInit;
+    abortSignal?: AbortSignal;
+    maxRedirects?: number;
+}): Promise<Response>;
+
 /**
  File content part of a prompt. It contains a file.
  */
@@ -2261,7 +2323,7 @@ declare interface GatewayGenerationInfoParams {
     id: string;
 }
 
-declare type GatewayImageModelId = 'bfl/flux-2-flex' | 'bfl/flux-2-klein-4b' | 'bfl/flux-2-klein-9b' | 'bfl/flux-2-max' | 'bfl/flux-2-pro' | 'bfl/flux-kontext-max' | 'bfl/flux-kontext-pro' | 'bfl/flux-pro-1.0-fill' | 'bfl/flux-pro-1.1' | 'bfl/flux-pro-1.1-ultra' | 'bytedance/seedream-4.0' | 'bytedance/seedream-4.5' | 'bytedance/seedream-5.0-lite' | 'google/imagen-4.0-fast-generate-001' | 'google/imagen-4.0-generate-001' | 'google/imagen-4.0-ultra-generate-001' | 'openai/gpt-image-1' | 'openai/gpt-image-1-mini' | 'openai/gpt-image-1.5' | 'openai/gpt-image-2' | 'prodia/flux-fast-schnell' | 'recraft/recraft-v2' | 'recraft/recraft-v3' | 'recraft/recraft-v4' | 'recraft/recraft-v4-pro' | 'xai/grok-imagine-image' | 'xai/grok-imagine-image-pro' | (string & {});
+declare type GatewayImageModelId = 'bfl/flux-2-flex' | 'bfl/flux-2-klein-4b' | 'bfl/flux-2-klein-9b' | 'bfl/flux-2-max' | 'bfl/flux-2-pro' | 'bfl/flux-kontext-max' | 'bfl/flux-kontext-pro' | 'bfl/flux-pro-1.0-fill' | 'bfl/flux-pro-1.1' | 'bfl/flux-pro-1.1-ultra' | 'bytedance/seedream-4.0' | 'bytedance/seedream-4.5' | 'bytedance/seedream-5.0-lite' | 'google/imagen-4.0-fast-generate-001' | 'google/imagen-4.0-generate-001' | 'google/imagen-4.0-ultra-generate-001' | 'openai/gpt-image-1' | 'openai/gpt-image-1-mini' | 'openai/gpt-image-1.5' | 'openai/gpt-image-2' | 'prodia/flux-fast-schnell' | 'quiverai/arrow-1.1' | 'recraft/recraft-v2' | 'recraft/recraft-v3' | 'recraft/recraft-v4' | 'recraft/recraft-v4-pro' | 'recraft/recraft-v4.1' | 'recraft/recraft-v4.1-pro' | 'recraft/recraft-v4.1-utility' | 'recraft/recraft-v4.1-utility-pro' | 'xai/grok-imagine-image' | (string & {});
 
 declare interface GatewayLanguageModelEntry {
     /**
@@ -2312,7 +2374,7 @@ declare interface GatewayLanguageModelEntry {
 
 declare type GatewayLanguageModelSpecification = Pick<LanguageModelV2, 'specificationVersion' | 'provider' | 'modelId'>;
 
-export declare type GatewayModelId = 'alibaba/qwen-3-14b' | 'alibaba/qwen-3-235b' | 'alibaba/qwen-3-30b' | 'alibaba/qwen-3-32b' | 'alibaba/qwen-3.6-max-preview' | 'alibaba/qwen3-235b-a22b-thinking' | 'alibaba/qwen3-coder' | 'alibaba/qwen3-coder-30b-a3b' | 'alibaba/qwen3-coder-next' | 'alibaba/qwen3-coder-plus' | 'alibaba/qwen3-max' | 'alibaba/qwen3-max-preview' | 'alibaba/qwen3-max-thinking' | 'alibaba/qwen3-next-80b-a3b-instruct' | 'alibaba/qwen3-next-80b-a3b-thinking' | 'alibaba/qwen3-vl-235b-a22b-instruct' | 'alibaba/qwen3-vl-instruct' | 'alibaba/qwen3-vl-thinking' | 'alibaba/qwen3.5-flash' | 'alibaba/qwen3.5-plus' | 'alibaba/qwen3.6-27b' | 'alibaba/qwen3.6-plus' | 'amazon/nova-2-lite' | 'amazon/nova-lite' | 'amazon/nova-micro' | 'amazon/nova-pro' | 'anthropic/claude-3-haiku' | 'anthropic/claude-3.5-haiku' | 'anthropic/claude-3.7-sonnet' | 'anthropic/claude-haiku-4.5' | 'anthropic/claude-opus-4' | 'anthropic/claude-opus-4.1' | 'anthropic/claude-opus-4.5' | 'anthropic/claude-opus-4.6' | 'anthropic/claude-opus-4.7' | 'anthropic/claude-sonnet-4' | 'anthropic/claude-sonnet-4.5' | 'anthropic/claude-sonnet-4.6' | 'arcee-ai/trinity-large-preview' | 'arcee-ai/trinity-large-thinking' | 'arcee-ai/trinity-mini' | 'bytedance/seed-1.6' | 'bytedance/seed-1.8' | 'cohere/command-a' | 'deepseek/deepseek-r1' | 'deepseek/deepseek-v3' | 'deepseek/deepseek-v3.1' | 'deepseek/deepseek-v3.1-terminus' | 'deepseek/deepseek-v3.2' | 'deepseek/deepseek-v3.2-thinking' | 'deepseek/deepseek-v4-flash' | 'deepseek/deepseek-v4-pro' | 'google/gemini-2.0-flash' | 'google/gemini-2.0-flash-lite' | 'google/gemini-2.5-flash' | 'google/gemini-2.5-flash-image' | 'google/gemini-2.5-flash-lite' | 'google/gemini-2.5-pro' | 'google/gemini-3-flash' | 'google/gemini-3-pro-image' | 'google/gemini-3-pro-preview' | 'google/gemini-3.1-flash-image-preview' | 'google/gemini-3.1-flash-lite-preview' | 'google/gemini-3.1-pro-preview' | 'google/gemma-4-26b-a4b-it' | 'google/gemma-4-31b-it' | 'inception/mercury-2' | 'inception/mercury-coder-small' | 'interfaze/interfaze-beta' | 'kwaipilot/kat-coder-pro-v1' | 'kwaipilot/kat-coder-pro-v2' | 'meituan/longcat-flash-chat' | 'meituan/longcat-flash-thinking-2601' | 'meta/llama-3.1-70b' | 'meta/llama-3.1-8b' | 'meta/llama-3.2-11b' | 'meta/llama-3.2-1b' | 'meta/llama-3.2-3b' | 'meta/llama-3.2-90b' | 'meta/llama-3.3-70b' | 'meta/llama-4-maverick' | 'meta/llama-4-scout' | 'minimax/minimax-m2' | 'minimax/minimax-m2.1' | 'minimax/minimax-m2.1-lightning' | 'minimax/minimax-m2.5' | 'minimax/minimax-m2.5-highspeed' | 'minimax/minimax-m2.7' | 'minimax/minimax-m2.7-highspeed' | 'mistral/codestral' | 'mistral/devstral-2' | 'mistral/devstral-small' | 'mistral/devstral-small-2' | 'mistral/magistral-medium' | 'mistral/magistral-small' | 'mistral/ministral-14b' | 'mistral/ministral-3b' | 'mistral/ministral-8b' | 'mistral/mistral-large-3' | 'mistral/mistral-medium' | 'mistral/mistral-nemo' | 'mistral/mistral-small' | 'mistral/pixtral-12b' | 'mistral/pixtral-large' | 'moonshotai/kimi-k2' | 'moonshotai/kimi-k2-thinking' | 'moonshotai/kimi-k2-thinking-turbo' | 'moonshotai/kimi-k2-turbo' | 'moonshotai/kimi-k2.5' | 'moonshotai/kimi-k2.6' | 'morph/morph-v3-fast' | 'morph/morph-v3-large' | 'nvidia/nemotron-3-nano-30b-a3b' | 'nvidia/nemotron-3-super-120b-a12b' | 'nvidia/nemotron-nano-12b-v2-vl' | 'nvidia/nemotron-nano-9b-v2' | 'openai/gpt-3.5-turbo' | 'openai/gpt-3.5-turbo-instruct' | 'openai/gpt-4-turbo' | 'openai/gpt-4.1' | 'openai/gpt-4.1-mini' | 'openai/gpt-4.1-nano' | 'openai/gpt-4o' | 'openai/gpt-4o-mini' | 'openai/gpt-4o-mini-search-preview' | 'openai/gpt-5' | 'openai/gpt-5-chat' | 'openai/gpt-5-codex' | 'openai/gpt-5-mini' | 'openai/gpt-5-nano' | 'openai/gpt-5-pro' | 'openai/gpt-5.1-codex' | 'openai/gpt-5.1-codex-max' | 'openai/gpt-5.1-codex-mini' | 'openai/gpt-5.1-instant' | 'openai/gpt-5.1-thinking' | 'openai/gpt-5.2' | 'openai/gpt-5.2-chat' | 'openai/gpt-5.2-codex' | 'openai/gpt-5.2-pro' | 'openai/gpt-5.3-chat' | 'openai/gpt-5.3-codex' | 'openai/gpt-5.4' | 'openai/gpt-5.4-mini' | 'openai/gpt-5.4-nano' | 'openai/gpt-5.4-pro' | 'openai/gpt-5.5' | 'openai/gpt-5.5-pro' | 'openai/gpt-oss-120b' | 'openai/gpt-oss-20b' | 'openai/gpt-oss-safeguard-20b' | 'openai/o1' | 'openai/o3' | 'openai/o3-deep-research' | 'openai/o3-mini' | 'openai/o3-pro' | 'openai/o4-mini' | 'perplexity/sonar' | 'perplexity/sonar-pro' | 'perplexity/sonar-reasoning-pro' | 'xai/grok-3' | 'xai/grok-3-fast' | 'xai/grok-3-mini' | 'xai/grok-3-mini-fast' | 'xai/grok-4' | 'xai/grok-4-fast-non-reasoning' | 'xai/grok-4-fast-reasoning' | 'xai/grok-4.1-fast-non-reasoning' | 'xai/grok-4.1-fast-reasoning' | 'xai/grok-4.20-multi-agent' | 'xai/grok-4.20-multi-agent-beta' | 'xai/grok-4.20-non-reasoning' | 'xai/grok-4.20-non-reasoning-beta' | 'xai/grok-4.20-reasoning' | 'xai/grok-4.20-reasoning-beta' | 'xai/grok-4.3' | 'xai/grok-code-fast-1' | 'xiaomi/mimo-v2-flash' | 'xiaomi/mimo-v2-pro' | 'xiaomi/mimo-v2.5' | 'xiaomi/mimo-v2.5-pro' | 'zai/glm-4.5' | 'zai/glm-4.5-air' | 'zai/glm-4.5v' | 'zai/glm-4.6' | 'zai/glm-4.6v' | 'zai/glm-4.6v-flash' | 'zai/glm-4.7' | 'zai/glm-4.7-flash' | 'zai/glm-4.7-flashx' | 'zai/glm-5' | 'zai/glm-5-turbo' | 'zai/glm-5.1' | 'zai/glm-5v-turbo' | (string & {});
+export declare type GatewayModelId = 'alibaba/qwen-3-14b' | 'alibaba/qwen-3-235b' | 'alibaba/qwen-3-30b' | 'alibaba/qwen-3-32b' | 'alibaba/qwen-3.6-max-preview' | 'alibaba/qwen3-235b-a22b-thinking' | 'alibaba/qwen3-coder' | 'alibaba/qwen3-coder-30b-a3b' | 'alibaba/qwen3-coder-next' | 'alibaba/qwen3-coder-plus' | 'alibaba/qwen3-max' | 'alibaba/qwen3-max-preview' | 'alibaba/qwen3-max-thinking' | 'alibaba/qwen3-next-80b-a3b-instruct' | 'alibaba/qwen3-next-80b-a3b-thinking' | 'alibaba/qwen3-vl-235b-a22b-instruct' | 'alibaba/qwen3-vl-instruct' | 'alibaba/qwen3-vl-thinking' | 'alibaba/qwen3.5-flash' | 'alibaba/qwen3.5-plus' | 'alibaba/qwen3.6-27b' | 'alibaba/qwen3.6-plus' | 'alibaba/qwen3.7-max' | 'alibaba/qwen3.7-plus' | 'amazon/nova-2-lite' | 'amazon/nova-lite' | 'amazon/nova-micro' | 'amazon/nova-pro' | 'anthropic/claude-3-haiku' | 'anthropic/claude-3.5-haiku' | 'anthropic/claude-fable-5' | 'anthropic/claude-haiku-4.5' | 'anthropic/claude-opus-4' | 'anthropic/claude-opus-4.1' | 'anthropic/claude-opus-4.5' | 'anthropic/claude-opus-4.6' | 'anthropic/claude-opus-4.7' | 'anthropic/claude-opus-4.8' | 'anthropic/claude-sonnet-4' | 'anthropic/claude-sonnet-4.5' | 'anthropic/claude-sonnet-4.6' | 'anthropic/claude-sonnet-5' | 'arcee-ai/trinity-large-preview' | 'arcee-ai/trinity-large-thinking' | 'arcee-ai/trinity-mini' | 'bytedance/seed-1.6' | 'bytedance/seed-1.8' | 'cohere/command-a' | 'deepseek/deepseek-r1' | 'deepseek/deepseek-v3' | 'deepseek/deepseek-v3.1' | 'deepseek/deepseek-v3.1-terminus' | 'deepseek/deepseek-v3.2' | 'deepseek/deepseek-v3.2-thinking' | 'deepseek/deepseek-v4-flash' | 'deepseek/deepseek-v4-pro' | 'google/gemini-2.5-flash' | 'google/gemini-2.5-flash-image' | 'google/gemini-2.5-flash-lite' | 'google/gemini-2.5-pro' | 'google/gemini-3-flash' | 'google/gemini-3-pro-image' | 'google/gemini-3-pro-preview' | 'google/gemini-3.1-flash-image' | 'google/gemini-3.1-flash-image-preview' | 'google/gemini-3.1-flash-lite' | 'google/gemini-3.1-flash-lite-image' | 'google/gemini-3.1-flash-lite-preview' | 'google/gemini-3.1-pro-preview' | 'google/gemini-3.5-flash' | 'google/gemma-4-26b-a4b-it' | 'google/gemma-4-31b-it' | 'inception/mercury-2' | 'inception/mercury-coder-small' | 'interfaze/interfaze-beta' | 'kwaipilot/kat-coder-pro-v1' | 'kwaipilot/kat-coder-pro-v2' | 'meituan/longcat-flash-chat' | 'meituan/longcat-flash-thinking-2601' | 'meta/llama-3.1-70b' | 'meta/llama-3.1-8b' | 'meta/llama-3.2-11b' | 'meta/llama-3.2-1b' | 'meta/llama-3.2-3b' | 'meta/llama-3.2-90b' | 'meta/llama-3.3-70b' | 'meta/llama-4-maverick' | 'meta/llama-4-scout' | 'minimax/minimax-m2' | 'minimax/minimax-m2.1' | 'minimax/minimax-m2.1-lightning' | 'minimax/minimax-m2.5' | 'minimax/minimax-m2.5-highspeed' | 'minimax/minimax-m2.7' | 'minimax/minimax-m2.7-highspeed' | 'minimax/minimax-m3' | 'mistral/codestral' | 'mistral/devstral-2' | 'mistral/devstral-small' | 'mistral/devstral-small-2' | 'mistral/magistral-medium' | 'mistral/magistral-small' | 'mistral/ministral-14b' | 'mistral/ministral-3b' | 'mistral/ministral-8b' | 'mistral/mistral-large-3' | 'mistral/mistral-medium' | 'mistral/mistral-medium-3.5' | 'mistral/mistral-nemo' | 'mistral/mistral-small' | 'mistral/pixtral-12b' | 'mistral/pixtral-large' | 'moonshotai/kimi-k2' | 'moonshotai/kimi-k2-thinking' | 'moonshotai/kimi-k2.5' | 'moonshotai/kimi-k2.6' | 'moonshotai/kimi-k2.7-code' | 'moonshotai/kimi-k2.7-code-highspeed' | 'morph/morph-v3-fast' | 'morph/morph-v3-large' | 'nvidia/nemotron-3-nano-30b-a3b' | 'nvidia/nemotron-3-super-120b-a12b' | 'nvidia/nemotron-3-ultra-550b-a55b' | 'nvidia/nemotron-nano-12b-v2-vl' | 'nvidia/nemotron-nano-9b-v2' | 'openai/gpt-3.5-turbo' | 'openai/gpt-3.5-turbo-instruct' | 'openai/gpt-4-turbo' | 'openai/gpt-4.1' | 'openai/gpt-4.1-mini' | 'openai/gpt-4.1-nano' | 'openai/gpt-4o' | 'openai/gpt-4o-mini' | 'openai/gpt-4o-mini-search-preview' | 'openai/gpt-5' | 'openai/gpt-5-chat' | 'openai/gpt-5-codex' | 'openai/gpt-5-mini' | 'openai/gpt-5-nano' | 'openai/gpt-5-pro' | 'openai/gpt-5.1-codex' | 'openai/gpt-5.1-codex-max' | 'openai/gpt-5.1-codex-mini' | 'openai/gpt-5.1-instant' | 'openai/gpt-5.1-thinking' | 'openai/gpt-5.2' | 'openai/gpt-5.2-chat' | 'openai/gpt-5.2-codex' | 'openai/gpt-5.2-pro' | 'openai/gpt-5.3-chat' | 'openai/gpt-5.3-codex' | 'openai/gpt-5.4' | 'openai/gpt-5.4-mini' | 'openai/gpt-5.4-nano' | 'openai/gpt-5.4-pro' | 'openai/gpt-5.5' | 'openai/gpt-5.5-pro' | 'openai/gpt-oss-120b' | 'openai/gpt-oss-20b' | 'openai/gpt-oss-safeguard-20b' | 'openai/o1' | 'openai/o3' | 'openai/o3-deep-research' | 'openai/o3-mini' | 'openai/o3-pro' | 'openai/o4-mini' | 'perplexity/sonar' | 'perplexity/sonar-pro' | 'perplexity/sonar-reasoning-pro' | 'sakana/fugu-ultra' | 'stepfun/step-3.5-flash' | 'stepfun/step-3.7-flash' | 'xai/grok-4.1-fast-non-reasoning' | 'xai/grok-4.1-fast-reasoning' | 'xai/grok-4.20-multi-agent' | 'xai/grok-4.20-multi-agent-beta' | 'xai/grok-4.20-non-reasoning' | 'xai/grok-4.20-non-reasoning-beta' | 'xai/grok-4.20-reasoning' | 'xai/grok-4.20-reasoning-beta' | 'xai/grok-4.3' | 'xai/grok-build-0.1' | 'xiaomi/mimo-v2-flash' | 'xiaomi/mimo-v2-pro' | 'xiaomi/mimo-v2.5' | 'xiaomi/mimo-v2.5-pro' | 'zai/glm-4.5' | 'zai/glm-4.5-air' | 'zai/glm-4.5v' | 'zai/glm-4.6' | 'zai/glm-4.6v' | 'zai/glm-4.6v-flash' | 'zai/glm-4.7' | 'zai/glm-4.7-flash' | 'zai/glm-4.7-flashx' | 'zai/glm-5' | 'zai/glm-5-turbo' | 'zai/glm-5.1' | 'zai/glm-5.2' | 'zai/glm-5.2-fast' | 'zai/glm-5v-turbo' | (string & {});
 
 declare interface GatewayProvider extends ProviderV2 {
     (modelId: GatewayModelId): LanguageModelV2;
@@ -3424,6 +3486,16 @@ export declare class InvalidToolInputError extends AISDKError {
 
 declare function isAbortError(error: unknown): error is Error;
 
+/**
+ * Returns `true` when running in a browser.
+ *
+ * Detection keys on the presence of a global `window`, matching the browser
+ * check used elsewhere in this package (see `getRuntimeEnvironmentUserAgent`)
+ * so the SDK has a single, consistent definition of "browser". Server runtimes
+ * (Node.js, Deno, Bun, edge/workers) do not define `window`.
+ */
+declare function isBrowserRuntime(globalThisAny?: any): boolean;
+
 /**
  * Check if a message part is a data part.
  */
@@ -3456,6 +3528,20 @@ declare function isParsableJson(input: string): boolean;
  */
 export declare function isReasoningUIPart(part: UIMessagePart<UIDataTypes, UITools>): part is ReasoningUIPart;
 
+/**
+ * Returns true when `url` has the same origin (scheme + host + port) as
+ * `baseUrl`.
+ *
+ * Used to decide whether provider credentials may be attached to a request to a
+ * URL taken from a provider response (e.g. a polling or media-download URL).
+ * Credentials must only be sent to the provider's own origin; a response that
+ * names a foreign host (a CDN, or an attacker-controlled host if the response
+ * is tampered with) must not receive the API key.
+ *
+ * Returns false if either value is not a valid absolute URL (fail-closed).
+ */
+declare function isSameOrigin(url: string, baseUrl: string): boolean;
+
 /**
  * Type guard to check if a message part is a text part.
  */
@@ -8714,6 +8800,10 @@ export declare const userModelMessageSchema: z.ZodType<UserModelMessage>;
  * Validates that a URL is safe to download from, blocking private/internal addresses
  * to prevent SSRF attacks.
  *
+ * Note: this performs string/literal-IP checks only. It does not resolve DNS, so a
+ * hostname that resolves to a private address is not blocked here (see callers, which
+ * should additionally constrain egress at the network layer when handling untrusted URLs).
+ *
  * @param url - The URL string to validate.
  * @throws DownloadError if the URL is unsafe.
  */

package/dist/_types/@internal_voice/dist/_types/@internal_core/dist/request-context/index.d.ts

@@ -129,6 +129,16 @@ declare class RequestContext<Values extends Record<string, any> | unknown = unkn
      * a clean JSON-safe dict for cross-context cycles.
      */
     private isSerializable;
+    /**
+     * Custom span serialization to prevent leaking internal state (like auth
+     * tokens stored in the private `registry` Map) into observability spans.
+     *
+     * `deepClean` in `@mastra/observability` calls this method before falling
+     * back to `Object.keys()` — which would walk the runtime-enumerable
+     * `registry` field and serialize its raw Map entries (including any
+     * bearer tokens) into exported spans.
+     */
+    serializeForSpan(): Record<string, unknown>;
     /**
      * Get all values as a typed object for destructuring.
      * Returns Record<string, any> when untyped, or the Values type when typed.

package/dist/docs/SKILL.md

@@ -3,7 +3,7 @@ name: mastra-voice-speechify
 description: Documentation for @mastra/voice-speechify. Use when working with @mastra/voice-speechify APIs, configuration, or implementation.
 metadata:
   package: "@mastra/voice-speechify"
-  version: "0.13.0"
+  version: "0.14.0"
 ---
 
 ## When to use
@@ -16,7 +16,6 @@ Read the individual reference documents for detailed explanations and code examp
 
 ### Docs
 
-- [Voice](references/docs-agents-adding-voice.md) - Learn how to add voice capabilities to your Mastra agents for text-to-speech and speech-to-text interactions.
 - [Voice in Mastra](references/docs-voice-overview.md) - Overview of voice capabilities in Mastra, including text-to-speech, speech-to-text, and real-time speech-to-speech interactions.
 
 ### Reference

package/dist/docs/assets/SOURCE_MAP.json

@@ -1,5 +1,5 @@
 {
-  "version": "0.13.0",
+  "version": "0.14.0",
   "package": "@mastra/voice-speechify",
   "exports": {},
   "modules": {}

package/dist/docs/references/docs-voice-overview.md

@@ -1,10 +1,12 @@
+> Discover all available pages from the documentation index: https://mastra.ai/llms.txt
+
 # Voice in Mastra
 
 Mastra's Voice system provides a unified interface for voice interactions, enabling text-to-speech (TTS), speech-to-text (STT), and real-time speech-to-speech (STS) capabilities in your applications.
 
-## Adding voice to agents
+## Add voice to agents
 
-To learn how to integrate voice capabilities into your agents, check out the [Adding Voice to Agents](https://mastra.ai/docs/agents/adding-voice) documentation. This section covers how to use both single and multiple voice providers, as well as real-time interactions.
+Pass a voice provider to an agent with the `voice` property. The same property supports text-to-speech (TTS), speech-to-text (STT), and real-time speech-to-speech (STS), depending on the provider you configure.
 
 ```typescript
 import { Agent } from '@mastra/core/agent'
@@ -1137,7 +1139,7 @@ const voiceAgent = new Agent({
 })
 ```
 
-### Using Multiple Voice Providers
+### Using multiple voice providers
 
 This example demonstrates how to create and use two different voice providers in Mastra: OpenAI for speech-to-text (STT) and PlayAI for text-to-speech (TTS).
 

package/dist/docs/references/reference-voice-speechify.md

@@ -1,3 +1,5 @@
+> Discover all available pages from the documentation index: https://mastra.ai/llms.txt
+
 # Speechify
 
 The Speechify voice implementation in Mastra provides text-to-speech capabilities using Speechify's API.
@@ -13,15 +15,15 @@ const voice = new SpeechifyVoice()
 // Initialize with custom configuration
 const voice = new SpeechifyVoice({
   speechModel: {
-    name: 'simba-english',
+    name: 'simba-3.2',
     apiKey: 'your-api-key',
   },
-  speaker: 'george', // Default voice
+  speaker: 'harper_32', // Default voice (simba-3.2 serves the curated Simba 3 voices only)
 })
 
 // Convert text to speech
 const audioStream = await voice.speak('Hello, world!', {
-  speaker: 'henry', // Override default voice
+  speaker: 'imogen_32', // Override default voice
 })
 ```
 
@@ -29,11 +31,11 @@ const audioStream = await voice.speak('Hello, world!', {
 
 **speechModel** (`SpeechifyConfig`): Configuration for text-to-speech functionality (Default: `{ name: 'simba-english' }`)
 
-**speechModel.name** (`VoiceModelName`): The Speechify model to use
+**speechModel.name** (`SpeechifyModel`): The Speechify model to use ('simba-3.2', 'simba-3.0', 'simba-english', or 'simba-multilingual')
 
 **speechModel.apiKey** (`string`): Speechify API key. Falls back to SPEECHIFY\_API\_KEY environment variable
 
-**speaker** (`SpeechifyVoiceId`): Default voice ID to use for speech synthesis (Default: `'george'`)
+**speaker** (`SpeechifyVoiceId`): Default voice ID to use for speech synthesis. The Simba 3 models serve a curated voice set only (harper\_32, imogen\_32, ...); the classic catalog voices (george, henry, ...) work with simba-english and simba-multilingual (Default: `'harper_32' for Simba 3 models, otherwise 'george'`)
 
 ## Methods
 
@@ -47,7 +49,7 @@ Converts text to speech using the configured speech model and voice.
 
 **options.speaker** (`string`): Override the default speaker for this request
 
-**options.model** (`VoiceModelName`): Override the default model for this request
+**options.model** (`SpeechifyModel`): Override the default model for this request
 
 Returns: `Promise<NodeJS.ReadableStream>`
 
@@ -71,5 +73,9 @@ This method isn't supported by Speechify and will throw an error. Speechify does
 
 - Speechify requires an API key for authentication
 - The default model is 'simba-english'
+- 'simba-3.2' is Speechify's latest streaming model with the lowest latency and richest expressivity, and the recommended model for English
+- 'simba-3.2' and 'simba-3.0' are currently English only; use 'simba-multilingual' for non-English or mixed-language input
+- 'simba-3.2' and 'simba-3.0' serve a curated voice set only: 'beatrice\_32', 'dominic\_32', 'edmund\_32', 'geffen\_32', 'harper\_32', 'hugh\_32', 'imogen\_32', 'wyatt\_32'. Classic catalog voices such as 'george' return an error on these models
+- The default speaker follows the configured model: 'harper\_32' for the Simba 3 models, otherwise 'george'
 - Speech-to-text functionality isn't supported
 - Additional audio stream options can be passed through the speak() method's options parameter

package/dist/index.cjs

@@ -265,6 +265,16 @@ var MastraVoice = class extends MastraBase {
 };
 
 // src/voices.ts
+var SIMBA_3_VOICES = [
+  "beatrice_32",
+  "dominic_32",
+  "edmund_32",
+  "geffen_32",
+  "harper_32",
+  "hugh_32",
+  "imogen_32",
+  "wyatt_32"
+];
 var SPEECHIFY_VOICES = [
   "henry",
   "bwyneth",
@@ -976,15 +986,19 @@ var SPEECHIFY_VOICES = [
 ];
 
 // src/index.ts
+var isSimba3Model = (model) => model === "simba-3.0" || model === "simba-3.2";
 var SpeechifyVoice = class extends MastraVoice {
   client;
   constructor({ speechModel, speaker } = {}) {
+    const modelName = speechModel?.name ?? "simba-english";
     super({
       speechModel: {
-        name: speechModel?.name ?? "simba-english",
+        name: modelName,
         apiKey: speechModel?.apiKey ?? process.env.SPEECHIFY_API_KEY
       },
-      speaker: speaker ?? "george"
+      // The Simba 3 models serve only the curated SIMBA_3_VOICES, so the
+      // default speaker has to follow the configured model.
+      speaker: speaker ?? (isSimba3Model(modelName) ? "harper_32" : "george")
     });
     const apiKey = speechModel?.apiKey ?? process.env.SPEECHIFY_API_KEY;
     if (!apiKey) {
@@ -993,7 +1007,7 @@ var SpeechifyVoice = class extends MastraVoice {
     this.client = new apiSdk.Speechify({ apiKey });
   }
   async getSpeakers() {
-    return SPEECHIFY_VOICES.map((voice) => ({
+    return [...SIMBA_3_VOICES, ...SPEECHIFY_VOICES].map((voice) => ({
       voiceId: voice,
       name: voice
     }));
@@ -1011,11 +1025,14 @@ var SpeechifyVoice = class extends MastraVoice {
   }
   async speak(input, options) {
     const text = typeof input === "string" ? input : await this.streamToString(input);
+    const { speaker, model, ...streamOptions } = options ?? {};
     const request = {
+      ...streamOptions,
       input: text,
-      model: options?.model || this.speechModel?.name,
-      voiceId: options?.speaker || this.speaker,
-      ...options
+      // The SDK sends `model` to the API verbatim, so casting the wider
+      // SpeechifyModel union into its narrower VoiceModelName is safe.
+      model: model || this.speechModel?.name,
+      voiceId: speaker || this.speaker
     };
     const webStream = await this.client.audioStream(request);
     const reader = webStream.getReader();