import {z} from 'zod' export type CreateApiV1AudioTranscriptionsArgs = {body: BodyInit}; export type CreateApiV1AudioTranscriptionsOptions = {baseUrl?: string | undefined, gatewayBearer?: string | undefined, fetch?: typeof fetch | undefined}; export const transcriptionLanguage = z.object({code: z.string()}); export const createTranscriptionResponseJson = z.object({text: z.string(), languages: z.array(transcriptionLanguage).optional(), logprobs: z.array(z.object({token: z.string().optional(), logprob: z.number().optional(), bytes: z.array(z.number()).optional()})).optional(), usage: z.union([z.object({type: z.literal("tokens"), input_tokens: z.number().int(), input_token_details: z.object({text_tokens: z.number().int().optional(), audio_tokens: z.number().int().optional()}).optional(), output_tokens: z.number().int(), total_tokens: z.number().int()}), z.object({type: z.literal("duration"), seconds: z.number()})]).optional()}); export const transcriptionDiarizedSegment = z.object({type: z.literal("transcript.text.segment"), id: z.string(), start: z.number(), end: z.number(), text: z.string(), speaker: z.string()}); export const createTranscriptionResponseDiarizedJson = z.object({task: z.literal("transcribe"), duration: z.number(), text: z.string(), segments: z.array(transcriptionDiarizedSegment), usage: z.discriminatedUnion("type", [z.object({type: z.literal("tokens"), input_tokens: z.number().int(), input_token_details: z.object({text_tokens: z.number().int().optional(), audio_tokens: z.number().int().optional()}).optional(), output_tokens: z.number().int(), total_tokens: z.number().int()}), z.object({type: z.literal("duration"), seconds: z.number()})]).optional()}); export const transcriptionWord = z.object({word: z.string(), start: z.number(), end: z.number()}); export const transcriptionSegment = z.object({id: z.number().int(), seek: z.number().int(), start: z.number(), end: z.number(), text: z.string(), tokens: z.array(z.number().int()), temperature: z.number(), avg_logprob: z.number(), compression_ratio: z.number(), no_speech_prob: z.number()}); export const transcriptTextUsageDuration = z.object({type: z.literal("duration"), seconds: z.number()}); export const createTranscriptionResponseVerboseJson = z.object({language: z.string(), duration: z.number(), text: z.string(), words: z.array(transcriptionWord).optional(), segments: z.array(transcriptionSegment).optional(), usage: transcriptTextUsageDuration.optional()}); export const createApiV1AudioTranscriptionsResponse = z.union([createTranscriptionResponseJson, createTranscriptionResponseDiarizedJson, createTranscriptionResponseVerboseJson]); export const validateResponse = async ( schema: { readonly '~standard': { readonly validate: (value: unknown) => ({ readonly value: Output; readonly issues?: undefined } | { readonly issues: ReadonlyArray<{ readonly message: string; readonly path?: ReadonlyArray | undefined }> }) | Promise<{ readonly value: Output; readonly issues?: undefined } | { readonly issues: ReadonlyArray<{ readonly message: string; readonly path?: ReadonlyArray | undefined }> }> } }, value: unknown ): Promise => { const result = await schema['~standard'].validate(value) if (result.issues !== undefined) { const detail = result.issues .map(({ message, path }) => path?.length ? `${path.map((segment) => String(typeof segment === 'object' ? segment.key : segment)).join('.')}: ${message}` : message ) .join('; ') throw new Error(`Response does not match the schema: ${detail}`, { cause: { issues: result.issues } }) } return result.value }; /** * /v1/audio/transcriptions * * POST /v1/audio/transcriptions */ export const createApiV1AudioTranscriptions = async (args: CreateApiV1AudioTranscriptionsArgs, options?: CreateApiV1AudioTranscriptionsOptions) => { const {baseUrl = "https://aihubmix.com", fetch: fetchFn = fetch, gatewayBearer} = options ?? {} const query = new URLSearchParams() const headers = new Headers() headers.set('accept', "application/json") if (gatewayBearer !== undefined) headers.set('authorization', `Bearer ${gatewayBearer}`) const url = `${baseUrl.replace(/\/+$/, '')}/v1/audio/transcriptions${query.size > 0 ? `?${query}` : ''}` const res = await fetchFn(url, {method: "POST", headers, body: args.body}) if (!res.ok) throw new Error(`Request failed with status ${res.status}`, {cause: {status: res.status, body: await res.text()}}) return validateResponse(createApiV1AudioTranscriptionsResponse, await res.json()) };