> ## Documentation Index > Fetch the complete documentation index at: https://docs.navtalk.ai/llms.txt > Use this file to discover all available pages before exploring further. # Create Voice > Create a new custom voice clone using either ElevenLabs or Cartesia provider. **Supported Providers:** - `elevenlabs`: ElevenLabs Instant Voice Cloning (IVC) - `cartesia`: Cartesia Voice Cloning ## Supported Languages The `language` parameter accepts the following language codes (supported by both providers): | Code | Language | Code | Language | Code | Language | |------|----------|------|----------|------|----------| | `en` | English | `zh` | Chinese | `ja` | Japanese | | `ko` | Korean | `fr` | French | `de` | German | | `es` | Spanish | `pt` | Portuguese | `hi` | Hindi | | `it` | Italian | `nl` | Dutch | `pl` | Polish | | `ru` | Russian | `sv` | Swedish | `tr` | Turkish | | `tl` | Tagalog | `bg` | Bulgarian | `ro` | Romanian | | `ar` | Arabic | `cs` | Czech | `el` | Greek | | `fi` | Finnish | `hr` | Croatian | `ms` | Malay | | `sk` | Slovak | `da` | Danish | `ta` | Tamil | | `uk` | Ukrainian | `hu` | Hungarian | `no` | Norwegian | | `vi` | Vietnamese | `bn` | Bengali | `th` | Thai | | `he` | Hebrew | `ka` | Georgian | `id` | Indonesian | | `te` | Telugu | `gu` | Gujarati | `kn` | Kannada | | `ml` | Malayalam | `mr` | Marathi | `pa` | Punjabi | **Note**: If `language` is not specified, the system will auto-detect the language from the audio file. ## Audio Requirements **Minimum Duration:** - **30 seconds minimum** (recommended: 60-120 seconds for best quality) - Shorter audio will result in error response **Supported Formats:** - MP3 (audio/mpeg) - WAV (audio/wav) **Quality Guidelines:** - Sample rate: 44.1 kHz recommended - Bit depth: 16-bit - Content: Clear speech, single speaker - Background: Minimal or no background noise - Background noise removal is automatically enabled (ElevenLabs) ## Next Steps After creating a voice: 1. Use the returned `voice` ID when creating avatars 2. The voice may take a few moments to generate preview 3. Voice is immediately available for text-to-speech synthesis ## OpenAPI ````yaml POST /api/open/v1/voice/{provider}/add openapi: 3.1.0 info: title: NavTalk API description: NavTalk API for AI avatar management, file uploads, and video generation version: 1.0.0 servers: - url: https://api.navtalk.ai description: NavTalk Production Environment Server - url: https://qaapi.navtalk.ai description: NavTalk QA Environment Server - url: https://devapi.navtalk.ai description: NavTalk Development Environment Server security: - navtalkLicenseAuth: [] paths: /api/open/v1/voice/{provider}/add: post: tags: - Voice summary: Create Voice description: >- Create a new custom voice clone using either ElevenLabs or Cartesia provider. **Supported Providers:** - `elevenlabs`: ElevenLabs Instant Voice Cloning (IVC) - `cartesia`: Cartesia Voice Cloning ## Supported Languages The `language` parameter accepts the following language codes (supported by both providers): | Code | Language | Code | Language | Code | Language | |------|----------|------|----------|------|----------| | `en` | English | `zh` | Chinese | `ja` | Japanese | | `ko` | Korean | `fr` | French | `de` | German | | `es` | Spanish | `pt` | Portuguese | `hi` | Hindi | | `it` | Italian | `nl` | Dutch | `pl` | Polish | | `ru` | Russian | `sv` | Swedish | `tr` | Turkish | | `tl` | Tagalog | `bg` | Bulgarian | `ro` | Romanian | | `ar` | Arabic | `cs` | Czech | `el` | Greek | | `fi` | Finnish | `hr` | Croatian | `ms` | Malay | | `sk` | Slovak | `da` | Danish | `ta` | Tamil | | `uk` | Ukrainian | `hu` | Hungarian | `no` | Norwegian | | `vi` | Vietnamese | `bn` | Bengali | `th` | Thai | | `he` | Hebrew | `ka` | Georgian | `id` | Indonesian | | `te` | Telugu | `gu` | Gujarati | `kn` | Kannada | | `ml` | Malayalam | `mr` | Marathi | `pa` | Punjabi | **Note**: If `language` is not specified, the system will auto-detect the language from the audio file. ## Audio Requirements **Minimum Duration:** - **30 seconds minimum** (recommended: 60-120 seconds for best quality) - Shorter audio will result in error response **Supported Formats:** - MP3 (audio/mpeg) - WAV (audio/wav) **Quality Guidelines:** - Sample rate: 44.1 kHz recommended - Bit depth: 16-bit - Content: Clear speech, single speaker - Background: Minimal or no background noise - Background noise removal is automatically enabled (ElevenLabs) ## Next Steps After creating a voice: 1. Use the returned `voice` ID when creating avatars 2. The voice may take a few moments to generate preview 3. Voice is immediately available for text-to-speech synthesis operationId: createVoice parameters: - name: provider in: path description: The voice provider type required: true schema: type: string enum: - elevenlabs - cartesia example: elevenlabs requestBody: required: true content: multipart/form-data: schema: $ref: '#/components/schemas/CreateVoiceRequest' encoding: file: contentType: audio/mpeg, audio/wav responses: '200': description: Voice created successfully content: application/json: schema: $ref: '#/components/schemas/CreateVoiceResponse' examples: elevenlabs: summary: ElevenLabs voice created value: code: 200 message: SUCCESS data: voice: J8K3LmN9PqR5StU7VwX1 name: My Custom Voice description: Custom voice created via API cartesia: summary: Cartesia voice created value: code: 200 message: SUCCESS data: voice: a0e99841-438c-4a64-b679-ae501e7d6091 name: My Cartesia Voice description: Custom Cartesia voice '400': description: Validation failed or creation error content: application/json: schema: $ref: '#/components/schemas/NavtalkErrorResponse' examples: providerNotFound: summary: Provider not found value: code: 400 message: Provider not found invalidProviderType: summary: Invalid provider type value: code: 400 message: >- Only ElevenLabs and Cartesia providers are supported for creating voices noPermission: summary: No permission to operate provider value: code: 400 message: You have no permission to operate this provider noFile: summary: Audio file required value: code: 400 message: Audio file is required durationTooShort: summary: Audio duration too short value: code: 400 message: >- Audio duration is too short: 15 seconds. Minimum required: 30 seconds (Recommended: 60-120 seconds) cannotParseAudio: summary: Cannot parse audio file value: code: 400 message: 无法解析音频时长：不支持的或损坏的音频文件 creationFailed: summary: Voice creation failed value: code: 400 message: 'Failed to create voice: API error details' security: - navtalkLicenseAuth: [] components: schemas: CreateVoiceRequest: required: - providerId - voiceName - file type: object properties: providerId: type: string description: The provider ID (ElevenLabs or Cartesia) example: b92a28fb2ea737457c7f13ba554759c0 voiceName: type: string description: Name for the new voice example: My Custom Voice language: type: string description: >- Optional language code for the voice. Supported by both ElevenLabs and Cartesia. See supported languages list in documentation. enum: - en - zh - ja - ko - fr - de - es - pt - hi - it - nl - pl - ru - sv - tr - tl - bg - ro - ar - cs - el - fi - hr - ms - sk - da - ta - uk - hu - 'no' - vi - bn - th - he - ka - id - te - gu - kn - ml - mr - pa example: zh description: type: string description: >- Optional description for the voice. Supported by both ElevenLabs and Cartesia. example: Custom cloned voice for my application file: type: string format: binary description: Audio file (MP3 or WAV, minimum 30 seconds) CreateVoiceResponse: type: object properties: code: type: integer example: 200 message: type: string example: SUCCESS data: type: object properties: voice: type: string description: The created voice ID (use as voice parameter) example: J8K3LmN9PqR5StU7VwX1 name: type: string description: The voice name example: My Custom Voice description: type: string description: Voice description example: Custom voice created via API NavtalkErrorResponse: type: object properties: code: type: integer example: 400 message: type: string example: Invalid license key securitySchemes: navtalkLicenseAuth: type: apiKey in: header name: license description: NavTalk API license key (e.g. sk_navtalk_...) ````