> ## Documentation Index > Fetch the complete documentation index at: https://developer.eka.care/llms.txt > Use this file to discover all available pages before exploring further. # Discovery > Public, unauthenticated endpoint that advertises the service's protocol version, supported authentication methods, audio capabilities, available models and supported languages. Clients should fetch this first and use the returned `models`, `capabilities.audio_formats` and `capabilities.upload_methods` to configure their session. The response is cacheable (`Cache-Control: max-age=10800`). The discovery endpoint is the **first call** a client should make. It is public (no authentication) and advertises everything a client needs to configure a session: the protocol version, supported authentication methods, audio capabilities, available models and supported languages. The response is cacheable — it is returned with `Cache-Control: max-age=10800` (3 hours). Fetch it once on startup and reuse it. ## What to do with the response | Field | Use it to… | | ----------------------------------------- | ---------------------------------------------------------------------------------------- | | `capabilities.audio_formats` | Pick a recording MIME type the server accepts. | | `capabilities.upload_methods` | Choose a valid `upload_type` (`chunked`, `single`, `stream`) for Create Session. | | `capabilities.max_chunk_duration_seconds` | Cap how long each uploaded chunk should be (≤20s recommended). | | `models[]` | Choose a `model` (`lite` / `pro`) based on languages, max session duration and features. | | `languages.supported` | Validate the `language_hint` / `transcript_language` you intend to send. | | `authentication.supported_methods` | Decide between API key and OIDC auth. | ## Response fields | Field | Type | Description | | -------------------- | --------- | ----------------------------------------------------------------- | | `protocol` | string | Always `medscribealliance`. | | `protocol_version` | string | Current protocol version (e.g. `0.1`). | | `supported_versions` | string\[] | All protocol versions the service supports. | | `service` | object | Service name, documentation URL, support email. | | `endpoints` | object | `base_url`, `webhooks_url`, `templates_url`. | | `authentication` | object | `supported_methods` and optional `oidc` config. | | `capabilities` | object | Audio formats, chunk limits, upload methods, delivery flags. | | `models` | object\[] | Available models with languages, limits and feature flags. | | `languages` | object | Supported language codes and whether auto-detection is available. | ## Next step Use the advertised models and capabilities to [create a session](/api-reference/health-ai/ekascribe/protocol/create-session). ## OpenAPI ````yaml GET /voice/v1/.well-known/medscribealliance openapi: 3.0.1 info: title: EkaScribe Protocol API (MedScribeAlliance v0.1) version: '0.1' description: > Session-based medical voice capture API implementing the MedScribeAlliance Protocol v0.1. Discover service capabilities, create a session, stream audio chunks, then end the session to trigger asynchronous transcription and structured-template extraction. servers: - url: https://api.eka.care description: Production server - url: https://api.dev.eka.care description: Development server security: [] paths: /voice/v1/.well-known/medscribealliance: get: tags: - protocol summary: Discovery description: > Public, unauthenticated endpoint that advertises the service's protocol version, supported authentication methods, audio capabilities, available models and supported languages. Clients should fetch this first and use the returned `models`, `capabilities.audio_formats` and `capabilities.upload_methods` to configure their session. The response is cacheable (`Cache-Control: max-age=10800`). responses: '200': description: Discovery document content: application/json: schema: $ref: '#/components/schemas/DiscoveryResponse' example: protocol: medscribealliance protocol_version: '0.1' supported_versions: - '0.1' service: name: Voice2Rx Medical Scribe Service documentation_url: https://developer.eka.care support_email: support@eka.care endpoints: base_url: https://api.eka.care/voice/v1 webhooks_url: https://api.eka.care/voice/v1/webhooks templates_url: https://api.eka.care/voice/api/v1/template authentication: supported_methods: - api_key - oidc oidc: issuer: https://accounts.eka.care/oauth2 authorization_endpoint: https://accounts.eka.care/oauth2/authorize token_endpoint: https://accounts.eka.care/oauth2/token scopes_supported: - openid - profile capabilities: audio_formats: - audio/webm;codecs=opus - audio/wav - audio/ogg - audio/mp4 - audio/m4a - audio/mp3 max_chunk_duration_seconds: 20 upload_methods: - chunked - single - stream webhook_delivery: true client_sdk_delivery: true models: - id: lite display_name: Lite languages: - en - hi max_session_duration_seconds: 600 response_speed: fast features: realtime_transcription: false speaker_diarization: false custom_templates: false - id: pro display_name: Professional languages: - en - hi - ta - te - bn - mr - gu - kn - ml - pa max_session_duration_seconds: 3600 response_speed: standard features: realtime_transcription: true speaker_diarization: true custom_templates: true languages: supported: - en - hi - ta - te - bn - mr - gu - kn - ml - pa auto_detection: true components: schemas: DiscoveryResponse: type: object properties: protocol: type: string example: medscribealliance protocol_version: type: string example: '0.1' supported_versions: type: array items: type: string service: type: object properties: name: type: string documentation_url: type: string support_email: type: string endpoints: type: object properties: base_url: type: string webhooks_url: type: string templates_url: type: string authentication: type: object properties: supported_methods: type: array items: type: string oidc: type: object properties: issuer: type: string authorization_endpoint: type: string token_endpoint: type: string scopes_supported: type: array items: type: string capabilities: type: object properties: audio_formats: type: array items: type: string max_chunk_duration_seconds: type: integer upload_methods: type: array items: type: string webhook_delivery: type: boolean client_sdk_delivery: type: boolean models: type: array items: type: object properties: id: type: string display_name: type: string languages: type: array items: type: string max_session_duration_seconds: type: integer response_speed: type: string features: type: object properties: realtime_transcription: type: boolean speaker_diarization: type: boolean custom_templates: type: boolean languages: type: object properties: supported: type: array items: type: string auto_detection: type: boolean ````