improvement(elevenlabs): wire stability and similarity_boost end-to-end (#4679)

* improvement(elevenlabs): wire stability and similarity_boost end-to-end

* fix(elevenlabs): guard NaN in voice settings and always send both knobs together

Author: waleedlatif1

apps/sim/app/api/tools/tts/route.ts

@@ -35,8 +35,17 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
     )
     if (!parsed.success) return parsed.response
 
-    const { text, voiceId, apiKey, modelId, workspaceId, workflowId, executionId } =
-      parsed.data.body
+    const {
+      text,
+      voiceId,
+      apiKey,
+      modelId,
+      stability,
+      similarityBoost,
+      workspaceId,
+      workflowId,
+      executionId,
+    } = parsed.data.body
 
     const voiceIdValidation = validateAlphanumericId(voiceId, 'voiceId', 255)
     if (!voiceIdValidation.isValid) {
@@ -57,6 +66,14 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
 
     const endpoint = `https://api.elevenlabs.io/v1/text-to-speech/${voiceId}`
 
+    const hasVoiceSetting = stability !== undefined || similarityBoost !== undefined
+    const voiceSettings = hasVoiceSetting
+      ? {
+          stability: stability ?? 0.5,
+          similarity_boost: similarityBoost ?? 0.75,
+        }
+      : undefined
+
     const response = await fetch(endpoint, {
       method: 'POST',
       headers: {
@@ -67,6 +84,7 @@ export const POST = withRouteHandler(async (request: NextRequest) => {
       body: JSON.stringify({
         text,
         model_id: modelId,
+        ...(voiceSettings ? { voice_settings: voiceSettings } : {}),
       }),
       signal: AbortSignal.timeout(DEFAULT_EXECUTION_TIMEOUT_MS),
     })

apps/sim/blocks/blocks/elevenlabs.ts

@@ -5,7 +5,7 @@ import type { ElevenLabsBlockResponse } from '@/tools/elevenlabs/types'
 export const ElevenLabsBlock: BlockConfig<ElevenLabsBlockResponse> = {
   type: 'elevenlabs',
   name: 'ElevenLabs',
-  description: 'Convert TTS using ElevenLabs',
+  description: 'Convert text to speech with ElevenLabs',
   authMode: AuthMode.ApiKey,
   longDescription: 'Integrate ElevenLabs into the workflow. Can convert text to speech.',
   docsLink: 'https://docs.sim.ai/tools/elevenlabs',
@@ -42,6 +42,21 @@ export const ElevenLabsBlock: BlockConfig<ElevenLabsBlockResponse> = {
         { label: 'eleven_flash_v2_5', id: 'eleven_flash_v2_5' },
         { label: 'eleven_v3', id: 'eleven_v3' },
       ],
+      value: () => 'eleven_monolingual_v1',
+    },
+    {
+      id: 'stability',
+      title: 'Stability',
+      type: 'short-input',
+      placeholder: '0.0 to 1.0 (e.g., 0.5)',
+      mode: 'advanced',
+    },
+    {
+      id: 'similarityBoost',
+      title: 'Similarity Boost',
+      type: 'short-input',
+      placeholder: '0.0 to 1.0 (e.g., 0.75)',
+      mode: 'advanced',
     },
     {
       id: 'apiKey',
@@ -57,19 +72,30 @@ export const ElevenLabsBlock: BlockConfig<ElevenLabsBlockResponse> = {
     access: ['elevenlabs_tts'],
     config: {
       tool: () => 'elevenlabs_tts',
-      params: (params) => ({
-        apiKey: params.apiKey,
-        text: params.text,
-        voiceId: params.voiceId,
-        modelId: params.modelId,
-      }),
+      params: (params) => {
+        const parseUnitInterval = (value: unknown): number | undefined => {
+          if (value === undefined || value === null || value === '') return undefined
+          const n = Number(value)
+          return Number.isFinite(n) ? n : undefined
+        }
+        return {
+          apiKey: params.apiKey,
+          text: params.text,
+          voiceId: params.voiceId,
+          modelId: params.modelId,
+          stability: parseUnitInterval(params.stability),
+          similarityBoost: parseUnitInterval(params.similarityBoost),
+        }
+      },
     },
   },
 
   inputs: {
     text: { type: 'string', description: 'Text to convert' },
     voiceId: { type: 'string', description: 'Voice identifier' },
     modelId: { type: 'string', description: 'Model identifier' },
+    stability: { type: 'number', description: 'Voice stability 0.0-1.0' },
+    similarityBoost: { type: 'number', description: 'Similarity boost 0.0-1.0' },
     apiKey: { type: 'string', description: 'ElevenLabs API key' },
   },
 

apps/sim/lib/api/contracts/tools/media/tts.ts

@@ -7,6 +7,8 @@ export const ttsToolBodySchema = z.object({
   voiceId: z.string({ error: 'Missing required parameters' }).min(1, 'Missing required parameters'),
   apiKey: z.string({ error: 'Missing required parameters' }).min(1, 'Missing required parameters'),
   modelId: z.string().optional().default('eleven_monolingual_v1'),
+  stability: z.coerce.number().min(0).max(1).optional(),
+  similarityBoost: z.coerce.number().min(0).max(1).optional(),
   workspaceId: z.string().optional(),
   workflowId: z.string().optional(),
   executionId: z.string().optional(),

apps/sim/tools/elevenlabs/tts.ts

@@ -4,7 +4,7 @@ import type { ToolConfig } from '@/tools/types'
 export const elevenLabsTtsTool: ToolConfig<ElevenLabsTtsParams, ElevenLabsTtsResponse> = {
   id: 'elevenlabs_tts',
   name: 'ElevenLabs TTS',
-  description: 'Convert TTS using ElevenLabs voices',
+  description: 'Convert text to speech using ElevenLabs voices',
   version: '1.0.0',
 
   params: {
@@ -34,7 +34,7 @@ export const elevenLabsTtsTool: ToolConfig<ElevenLabsTtsParams, ElevenLabsTtsRes
       description:
         'Voice stability setting from 0.0 to 1.0 (e.g., 0.5 for balanced, 0.75 for more stable). Higher values produce more consistent output',
     },
-    similarity: {
+    similarityBoost: {
       type: 'number',
       required: false,
       visibility: 'user-or-llm',
@@ -52,7 +52,7 @@ export const elevenLabsTtsTool: ToolConfig<ElevenLabsTtsParams, ElevenLabsTtsRes
   request: {
     url: '/api/tools/tts',
     method: 'POST',
-    headers: (params) => ({
+    headers: () => ({
       'Content-Type': 'application/json',
     }),
     body: (
@@ -65,7 +65,7 @@ export const elevenLabsTtsTool: ToolConfig<ElevenLabsTtsParams, ElevenLabsTtsRes
       voiceId: params.voiceId,
       modelId: params.modelId || 'eleven_monolingual_v1',
       stability: params.stability,
-      similarity: params.similarity,
+      similarityBoost: params.similarityBoost,
       workspaceId: params._context?.workspaceId,
       workflowId: params._context?.workflowId,
       executionId: params._context?.executionId,

apps/sim/tools/elevenlabs/types.ts

@@ -7,7 +7,7 @@ export interface ElevenLabsTtsParams {
   voiceId: string
   modelId?: string
   stability?: number
-  similarity?: number
+  similarityBoost?: number
 }
 
 export interface ElevenLabsTtsResponse extends ToolResponse {
@@ -17,9 +17,4 @@ export interface ElevenLabsTtsResponse extends ToolResponse {
   }
 }
 
-export interface ElevenLabsBlockResponse extends ToolResponse {
-  output: {
-    audioUrl: string
-    audioFile?: UserFile
-  }
-}
+export type ElevenLabsBlockResponse = ElevenLabsTtsResponse