> ## Documentation Index > Fetch the complete documentation index at: https://docs.aurigin.ai/llms.txt > Use this file to discover all available pages before exploring further. # AI-voice detection > Detect whether audio clips are human (`bonafide`) or AI-generated (`spoofed`) and get segment-level confidence scores. **Designed for quick checks** - Supported formats: WAV, MP3, AAC, FLAC, OGG, M4A, MP4, MOV, AVI, MKV **Two ways to send audio** 1. `multipart/form-data` upload with a `file` 2. `application/json` payload with a `presigned_url` The API automatically analyzes the audio in ~5-second segments and returns both a global verdict and detailed per-segment results. ## OpenAPI ````yaml /api-reference/v1/openapi.json post /predict openapi: 3.0.3 info: title: Aurigin API version: 1.0.0 description: Updated API with new response structure - v1.0.0 servers: - url: https://api.aurigin.ai/v1 security: [] paths: /predict: post: tags: - v1 summary: AI-voice detection description: >- Detect whether audio clips are human (`bonafide`) or AI-generated (`spoofed`) and get segment-level confidence scores. **Designed for quick checks** - Supported formats: WAV, MP3, AAC, FLAC, OGG, M4A, MP4, MOV, AVI, MKV **Two ways to send audio** 1. `multipart/form-data` upload with a `file` 2. `application/json` payload with a `presigned_url` The API automatically analyzes the audio in ~5-second segments and returns both a global verdict and detailed per-segment results. operationId: predictV1 requestBody: required: true description: >- Provide either a direct file upload or a presigned URL. Include `device` to tag the caller and optionally choose a `model`. content: multipart/form-data: schema: type: object properties: file: type: string format: binary device: type: string description: Optional device type making the request enum: - macos - windows - web_app - api model: type: string description: >- Optional model version to run. Defaults to the latest available model (`apollo-4-5-2026-03-26`). `apollo-4-2026-01-16` remains available for backwards compatibility. enum: - apollo-4-5-2026-03-26 - apollo-4-2026-01-16 nullable: true prediction_id: type: string description: Optional custom prediction identifier silence_threshold: type: number description: >- Maximum silence percentage allowed per segment before it is skipped (0.0-100.0). Segments where silence exceeds this threshold are excluded from model inference. Lower values are stricter (skip more segments); higher values are more permissive. Set to `100` to analyze virtually all segments regardless of silence. default: 80 minimum: 0 maximum: 100 threshold: type: number description: >- Decision threshold for classifying audio as bonafide or spoofed (0.0-1.0). Lower values bias toward higher spoof detection; higher values reduce false positives. Confidence scores are derived from how far the raw score is from this threshold. default: 0.5 minimum: 0 maximum: 1 required: - file example: file: (binary audio file) device: api model: apollo-4-5-2026-03-26 threshold: 0.5 application/json: schema: $ref: '#/components/schemas/PredictRequest' example: presigned_url: https://example.com/presigned.wav device: api model: apollo-4-5-2026-03-26 threshold: 0.5 responses: '200': description: Global verdict plus per-segment predictions and confidence scores. content: application/json: schema: $ref: '#/components/schemas/PredictResponse' examples: spoofed: summary: Spoofed audio detected value: prediction_id: pred_9b6ff057a7f7 global: score: 0.975 confidence: 0.95 result: spoofed reason: null segments: - start: 0 end: 5 scores: - 0.98 confidence: 0.96 result: spoofed - start: 5 end: 10 scores: - 0.97 confidence: 0.94 result: spoofed model: apollo-4-5-2026-03-26 processing_time: 1.23 audio_duration: 10 warnings: [] bonafide: summary: Authentic audio value: prediction_id: pred_abc123def456 global: score: 0.04 confidence: 0.92 result: bonafide reason: null segments: - start: 0 end: 5 scores: - 0.045 confidence: 0.91 result: bonafide - start: 5 end: 8.5 scores: - 0.035 confidence: 0.93 result: bonafide model: apollo-4-5-2026-03-26 processing_time: 0.87 audio_duration: 8.5 warnings: [] '400': description: Validation error (for example, unsupported format). content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: validation_error message: Unsupported audio format status: 400 correlation_id: a1b2c3d4-e5f6-7890-abcd-ef1234567890 '401': description: Missing or invalid API key. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: unauthorized message: Invalid or inactive API key status: 401 correlation_id: 8b738992-c4eb-4f19-870a-900e6830d147 '413': description: Payload too large (over 5 MB). content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: payload_too_large message: Request body exceeds maximum allowed size of 5MB status: 413 correlation_id: 7f629881-a3db-5e28-761b-811f7940c258 '422': description: The audio could not be processed (for example, corrupted media). content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: unsupported_format message: Provided audio format is not supported status: 422 correlation_id: 48cd2c61-9cbe-4761-8a8f-f65aabbd22b9 '500': description: Unexpected server error while running the prediction. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: processing_failed message: Internal error occurred while analyzing audio status: 500 correlation_id: 1dd6c5a3-bda8-4de6-9171-e14d8d31c88e '503': description: Service temporarily unavailable — retry with backoff. content: application/json: schema: $ref: '#/components/schemas/ErrorResponse' example: error: service_unavailable message: >- Service temporarily unavailable, retry with exponential backoff status: 503 correlation_id: 5f4f28d6-78f4-4202-a49f-96ce7390d5d1 security: - api_key: [] externalDocs: description: Guided walkthrough and code samples url: https://docs.aurigin.ai/api-reference/v1/endpoint/predict components: schemas: PredictRequest: type: object properties: user_id: type: string description: Optional user identifier presigned_url: type: string description: Presigned URL to the audio file device: type: string description: Optional device type making the request enum: - macos - windows - web_app - api model: type: string description: >- Optional model version to run. Defaults to the latest available model (`apollo-4-5-2026-03-26`). enum: - apollo-4-5-2026-03-26 - apollo-4-2026-01-16 silence_threshold: type: number description: >- Maximum silence percentage allowed per segment before it is skipped (0.0-100.0). Segments where silence exceeds this threshold are excluded from model inference. Lower values are stricter (skip more segments); higher values are more permissive. Set to `100` to analyze virtually all segments regardless of silence. default: 80 minimum: 0 maximum: 100 threshold: type: number description: >- Decision threshold for classifying audio as bonafide or spoofed (0.0-1.0). Lower values bias toward higher spoof detection; higher values reduce false positives. Confidence scores are derived from how far the raw score is from this threshold. default: 0.5 minimum: 0 maximum: 1 required: - presigned_url PredictResponse: type: object properties: prediction_id: type: string description: Unique identifier for this prediction global: type: object properties: score: type: number format: float description: >- Raw probabilistic score (0.0-1.0). 0 indicates bonafide, 1 indicates spoofed. minimum: 0 maximum: 1 confidence: type: number format: float description: >- Overall confidence score for the entire audio file (0.0-1.0), derived from how far the score is from the decision threshold result: type: string enum: - bonafide - spoofed - partially_spoofed description: Overall classification result for the entire audio file reason: type: string nullable: true description: Optional reason or explanation for the result segments: type: array items: type: object properties: start: type: number format: float description: Start time of the segment in seconds end: type: number format: float description: End time of the segment in seconds scores: type: array items: type: number format: float description: >- Array of raw probabilistic scores for the sub-segments aggregated into this segment (each 0.0-1.0). 0 indicates bonafide, 1 indicates spoofed. Multiple values appear when several sub-segments are combined into one segment. confidence: type: number format: float description: >- Confidence score for this segment (0.0-1.0), derived from how far the score is from the decision threshold result: type: string enum: - bonafide - spoofed description: Classification result for this segment description: >- Array of segment-level predictions. Each segment represents a 5-second window of the audio. model: type: string description: Model version used for prediction processing_time: type: number format: float description: Time taken to process the audio in seconds audio_duration: type: number format: float description: Duration of the audio file in seconds warnings: type: array items: type: string description: Array of warning messages, if any example: prediction_id: pred_c3f91c4e10f3 global: score: 0.49742 confidence: 0.34647911317705393 result: partially_spoofed reason: null segments: - start: 0 end: 5 scores: - 0.00001 confidence: 0.000006612953711737646 result: bonafide - start: 5 end: 10 scores: - 0.99484 confidence: 0.9948440790176392 result: spoofed model: apollo-4-5-2026-03-26 processing_time: 11.55466890335083 audio_duration: 41.51 warnings: [] ErrorResponse: type: object properties: error: type: string description: Machine-readable error identifier example: validation_error message: type: string description: Human-readable explanation of the error example: Unsupported audio format status: type: integer description: HTTP status code associated with the error example: 400 correlation_id: type: string description: Unique identifier for tracing failed requests example: a1b2c3d4-e5f6-7890-abcd-ef1234567890 securitySchemes: api_key: type: apiKey in: header name: x-api-key ````