{ "openapi": "3.0.3", "info": { "title": "Oruk Qwen3-ASR API", "version": "1.0.0", "description": "Multilingual file transcription with Qwen/Qwen3-ASR-1.7B, developed by Qwen / Alibaba and hosted by Oruk on NVIDIA L4. One file per request; maximum 60 seconds decoded duration and 25 MiB. Returns complete text; streaming, timestamps and diarization are not enabled. Direct origin access requires a separately issued Oruk credential. Public paid access is available on RapidAPI and Poe; platform pricing/terms are separate. HF Inference Providers approval remains pending; the SDK proposal is Ready for review in PR https://github.com/huggingface/huggingface.js/pull/2680.", "contact": { "name": "Oruk", "url": "https://oruk.ai", "email": "nathan@oruk.ai" }, "x-model-id": "Qwen/Qwen3-ASR-1.7B", "x-model-developer": "Qwen / Alibaba", "x-model-license": "Apache-2.0", "x-api-operator": "Oruk, Inc.", "x-public-access-links": { "rapidapi": "https://rapidapi.com/NathanRoll/api/oruk-speech-api", "poe": "https://poe.com/Oruk-Qwen3-ASR" } }, "servers": [ { "url": "https://asia.asr.oruk.ai", "description": "Explicit Oruk regional ASR origin; retain region for billing and idempotency." } ], "tags": [ { "name": "Speech recognition", "description": "Complete-file audio transcription." } ], "externalDocs": { "description": "Production API documentation", "url": "https://qwen3-asr.oruk.ai/docs" }, "paths": { "/v1/audio/transcriptions": { "post": { "tags": [ "Speech recognition" ], "summary": "Transcribe one audio file", "operationId": "transcribeAudio", "description": "Upload binary audio using multipart/form-data. Let the HTTP client set Content-Type and its multipart boundary. Unsupported or duplicate fields are rejected. This route does not fetch audio URLs.", "requestBody": { "required": true, "content": { "multipart/form-data": { "schema": { "type": "object", "required": [ "model", "file" ], "additionalProperties": false, "properties": { "model": { "type": "string", "enum": [ "Qwen/Qwen3-ASR-1.7B" ], "description": "Exactly Qwen/Qwen3-ASR-1.7B." }, "file": { "type": "string", "format": "binary", "description": "One binary audio file, maximum 25 MiB (26,214,400 bytes) and 60 seconds decoded duration. Audio URLs and JSON upload wrappers are not supported by the general origin route." }, "language": { "type": "string", "pattern": "^[a-z]{2,3}$", "description": "Optional lowercase ISO language code; omit for automatic detection." }, "prompt": { "type": "string", "maxLength": 1000, "description": "Optional transcription context. Control tokens such as <| and are rejected." }, "response_format": { "type": "string", "enum": [ "json", "text" ], "default": "json", "description": "json returns an object with text; text returns a plain transcript." }, "temperature": { "type": "number", "enum": [ 0 ], "default": 0, "description": "Only deterministic temperature zero is enabled." }, "stream": { "type": "boolean", "enum": [ false ], "default": false, "description": "Must be false. This deployment returns a complete file transcript; streaming is not enabled." } } } } } }, "responses": { "200": { "description": "Complete transcript. The response media type follows response_format.", "headers": { "Inference-Id": { "description": "Request UUID generated by the Oruk gateway.", "schema": { "type": "string", "format": "uuid" } } }, "content": { "application/json": { "schema": { "type": "object", "required": [ "text" ], "properties": { "text": { "type": "string" } }, "additionalProperties": false }, "example": { "text": "Your transcript appears here." } }, "text/plain": { "schema": { "type": "string" }, "example": "Your transcript appears here." } } }, "400": { "description": "Invalid audio, unsupported model/options, duplicate fields or invalid parameters.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "401": { "description": "Missing or invalid authentication.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "409": { "description": "Idempotency key has already been used; the original transcript is not retained.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "413": { "description": "Upload exceeds 25 MiB or decoded audio exceeds 60 seconds.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "429": { "description": "Inference capacity is busy; retry with backoff.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "502": { "description": "Model engine unavailable or returned an invalid response.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } }, "504": { "description": "Model engine timed out.", "content": { "application/json": { "schema": { "type": "object", "additionalProperties": true } } } } }, "x-max-audio-seconds": 60, "x-max-file-bytes": 26214400 } }, "/v1/models": { "get": { "summary": "List the model served by this origin", "operationId": "listModels", "description": "Requires the same Oruk bearer credential as transcription. Returns the current model, decoded-duration limit and configured direct-origin rate.", "responses": { "200": { "description": "Available model catalog.", "headers": { "Inference-Id": { "schema": { "type": "string", "format": "uuid" } } }, "content": { "application/json": { "schema": { "type": "object", "required": [ "object", "data" ], "additionalProperties": false, "properties": { "object": { "type": "string", "enum": [ "list" ] }, "data": { "type": "array", "items": { "type": "object", "required": [ "id", "object", "created", "owned_by", "served_by", "task", "max_audio_seconds", "pricing" ], "additionalProperties": false, "properties": { "id": { "type": "string", "enum": [ "Qwen/Qwen3-ASR-1.7B" ] }, "object": { "type": "string", "enum": [ "model" ] }, "created": { "type": "integer", "format": "int64", "description": "Model creation time as a Unix timestamp." }, "owned_by": { "type": "string", "enum": [ "Qwen" ] }, "served_by": { "type": "string", "enum": [ "Oruk" ] }, "task": { "type": "string", "enum": [ "automatic-speech-recognition" ] }, "max_audio_seconds": { "type": "integer", "minimum": 1 }, "pricing": { "type": "object", "description": "Configured direct-origin rate. Marketplace charges are defined separately by each platform.", "required": [ "unit", "nano_usd", "usd" ], "additionalProperties": false, "properties": { "unit": { "type": "string", "enum": [ "audio_minute" ] }, "nano_usd": { "type": "integer", "format": "int64", "minimum": 0 }, "usd": { "type": "string", "pattern": "^\\d+\\.\\d{9}$" } } } } } } } } } } }, "401": { "description": "Missing or invalid Oruk bearer credential." } } } } }, "components": { "securitySchemes": { "BearerAuth": { "type": "http", "scheme": "bearer", "description": "Supply a separately issued Oruk key. Marketplace consumers use their marketplace key at the marketplace URL; backend credentials are configured privately by the distributor." } } }, "security": [ { "BearerAuth": [] } ] }