# Score a transcript or a recording ## OpenAPI Specification ```yaml openapi: 3.0.1 info: title: '' description: '' version: 1.0.0 paths: /api/voice/scan: post: summary: Score a transcript or a recording deprecated: false description: >- Classify one call after the fact. Send a transcript, or audio as base64 or as a multipart file named `audio`. Exactly one of transcript or audio is required. Costs 2 credits per request plus 1 credit per started minute of uploaded audio. A request rejected with 400 costs nothing. The body limit is 25 MB. The request times out after 90 seconds. With `report=true` and `from` present, a score of 0.8 or more files the caller and every callback number as a complaint. `none`, `Advertising`, `Surveys`, `Hang Up`, and `No Subject Provided` are never filed. `Other` needs 0.85. Multipart is also accepted: a file field named `audio` plus the same fields as form values. operationId: voiceScan tags: - Spam protection/Voice firewall/REST parameters: [] requestBody: content: application/json: schema: type: object description: Exactly one of `transcript` or `audio_base64` is required. properties: transcript: type: string description: Text of the call. audio_base64: type: string description: Base64 audio. encoding: type: string enum: - wav - pcm16 - mulaw default: wav description: >- `wav` needs a RIFF header. `pcm16` is 16-bit little-endian and needs `sample_rate`. `mulaw` defaults to 8000 Hz. sample_rate: type: integer minimum: 8000 maximum: 48000 description: Required for `pcm16`. language: type: string description: BCP-47 hint for transcription. from: type: string description: >- Caller number in E.164 with a leading `+`. Required for `report`. to: type: string description: Called number in E.164. report: type: boolean default: false description: File the caller when the score is 0.8 or more. responses: '200': description: The verdict, the intel, and the transcript. content: application/json: schema: type: object required: - status - session_id - verdict - intel - transcript - reported - reported_numbers properties: status: type: string enum: - success session_id: type: string format: uuid verdict: type: object required: - score - level - category - tactics - signals - source properties: score: type: number minimum: 0 maximum: 1 description: Probability that the caller is a scammer. level: type: string enum: - clean - suspicious - likely_scam - scam description: >- clean below 0.3. suspicious 0.3 to 0.59. likely_scam 0.6 to 0.84. scam 0.85 and above. category: type: string description: >- One exact CallerAPI complaint subject, or `none`. See GET /api/voice/categories. tactics: type: array items: type: string enum: - urgency - secrecy - authority_impersonation - remote_access - untraceable_payment - pii_harvest - otp_request - threat - prize_lure - family_fear - refund_bait - callback_pressure signals: type: array items: type: string description: >- Short pieces of evidence. Above a score of 0.6 they quote the caller. summary: type: string description: >- Two sentences at most. Present when a model produced the verdict. source: type: string enum: - heuristic - llm - combined - stored intel: type: object description: >- Facts the caller exposed. Every field is omitted when empty. properties: names: type: array items: type: string organizations: type: array items: type: string callback_numbers: type: array items: type: string description: E.164. Filed with the caller when `report` is on. emails: type: array items: type: string urls: type: array items: type: string crypto_wallets: type: array items: type: string bank_accounts: type: array items: type: string card_numbers: type: array items: type: string gift_card_brands: type: array items: type: string payment_apps: type: array items: type: string remote_tools: type: array items: type: string remote_codes: type: array items: type: string case_numbers: type: array items: type: string amounts: type: array items: type: string addresses: type: array items: type: string scripts: type: array items: type: string description: Verbatim lines of the pitch. dtmf: type: string description: Every digit pressed, in order. transcript: type: string reported: type: boolean reported_numbers: type: integer description: 'Complaints filed: the caller plus every callback number.' example: status: success session_id: 3f0c1a2e-8b1d-4f7e-9c2a-1d2e3f4a5b6c verdict: score: 0.93 level: scam category: >- Calls pretending to be government, businesses, or family and friends tactics: - authority_impersonation - threat - untraceable_payment - urgency signals: - invokes Social Security - asks for gift cards - threatens arrest or legal action summary: >- Caller impersonates Social Security and demands gift cards under threat of arrest. source: combined intel: {} transcript: >- This is officer Smith from Social Security. Pay with gift cards today or you will be arrested. reported: false reported_numbers: 0 headers: {} x-apidog-name: The verdict, the intel, and the transcript '400': description: Bad input. The message says which field. content: application/json: schema: type: object required: - status - message properties: status: type: string enum: - error message: type: string example: status: error message: 'forward is required: the E.164 number the call should ring' headers: {} x-apidog-name: Bad input '401': description: Missing or wrong API key. content: application/json: schema: type: object required: - status - message properties: status: type: string enum: - error message: type: string example: status: error message: Unauthorized headers: {} x-apidog-name: Missing or wrong API key '402': description: Insufficient credits. content: application/json: schema: type: object required: - status - message properties: status: type: string enum: - error message: type: string example: status: error message: Insufficient credits headers: {} x-apidog-name: Insufficient credits '502': description: Transcription failed. content: application/json: schema: type: object required: - status - message properties: status: type: string enum: - error message: type: string example: status: error message: Transcription failed headers: {} x-apidog-name: Transcription failed '503': description: Audio was sent but speech recognition is off on this deployment. content: application/json: schema: type: object required: - status - message properties: status: type: string enum: - error message: type: string example: status: error message: >- No speech-to-text provider is configured; send a transcript instead headers: {} x-apidog-name: Audio was sent but speech recognition is off on this deploym security: - CallerAPI Key: [] x-apidog: schemeGroups: - id: 7Wgpz7XqJ5wdLyzsKr91e schemeIds: - CallerAPI Key required: true use: id: 7Wgpz7XqJ5wdLyzsKr91e x-apidog-folder: Spam protection/Voice firewall/REST x-apidog-status: released x-run-in-apidog: https://app.apidog.com/web/project/1027392/apis/api-43843796-run components: schemas: {} securitySchemes: CallerAPI Key: type: apikey in: header name: x-auth ApiKeyAuth: type: apikey in: header name: X-Auth description: CallerAPI API key (UUID). servers: - url: https://api.callerapi.com description: Prod Env security: - CallerAPI Key: [] x-apidog: schemeGroups: - id: 7Wgpz7XqJ5wdLyzsKr91e schemeIds: - CallerAPI Key required: true use: id: 7Wgpz7XqJ5wdLyzsKr91e ```