From 68ee44906a85ff1bb6ce10671538ca027406a8f3 Mon Sep 17 00:00:00 2001 From: Rishabh Bhargava Date: Thu, 7 May 2026 15:37:34 -0700 Subject: [PATCH] MLE-5325: docs(audio): drop task field from verbose_json schemas The task field was specified as required in both AudioTranscriptionVerboseJsonResponse and AudioTranslationVerboseJsonResponse but was never built by inference-pop or tvllm. The endpoint URL already discriminates the task (/audio/transcriptions vs /audio/translations), so the field carries no information the caller doesn't already have. Dropping from the spec rather than implementing it; this aligns the documented contract with what the API actually returns today. Note: anyone using strict OpenAI-compat SDKs (openai-python TranscriptionVerbose) that expects `task` will need to skip strict validation. This was the existing state already, since the field was never returned. --- openapi.yaml | 16 ---------------- 1 file changed, 16 deletions(-) diff --git a/openapi.yaml b/openapi.yaml index e8dd678..8b0b61c 100644 --- a/openapi.yaml +++ b/openapi.yaml @@ -10244,19 +10244,11 @@ components: AudioTranscriptionVerboseJsonResponse: type: object required: - - task - language - duration - text - segments properties: - task: - type: string - description: The task performed - enum: - - transcribe - - translate - example: transcribe language: type: string description: The language of the audio @@ -10460,19 +10452,11 @@ components: AudioTranslationVerboseJsonResponse: type: object required: - - task - language - duration - text - segments properties: - task: - type: string - description: The task performed - enum: - - transcribe - - translate - example: translate language: type: string description: The target language of the translation