OpenAI

[English]

openai 组件是 OpenAI-ESP32 Arduino 库的移植版本，它简化了在 ESP-IDF 上调用 OpenAI API 的过程。该库广泛支持大部分 OpenAI API 的功能，但不包括 files 和 fine-tunes。更多详情，请参考 OpenAI 的 API 参考文档。

FAQ

使用 audioTranscription->file 报错 Failed to allocate request buffer!

该错误表明动态内存分配失败，通常是由于可用内存不足导致的。所需的内存大小取决于当前剩余的 RAM 或 PSRAM 大小。

解决方法

开启 PSRAM：如果设备支持 PSRAM，可以开启 PSRAM ，以增加可用内存，并打开宏 CONFIG_SPIRAM_USE_MALLOC 以使用 PSRAM。
减少音频的数据量:减少音频的数据量，例如减少音频的采样率、音频的长度等。

API 参考

Header File

components/openai/include/OpenAI.h

Functions

OpenAI_t *OpenAICreate(const char *api_key)

Create an OpenAI object.

参数: api_key – The key of openai
返回: OpenAI_t* The OpenAI object

void OpenAIDelete(OpenAI_t *oai)

Clear the OpenAI object and release resources.

参数: oai – The OpenAI object

void OpenAIChangeBaseURL(OpenAI_t *oai, const char *baseURL): Modify the Base URL of the OpenAI object.

Structures

struct OpenAI_EmbeddingData_t

Struct for Embedding data.

Public Members

uint32_t len: Length of the data

double *data: Pointer to the data

struct OpenAI_EmbeddingResponse

To get an embedding, send your text string to the embeddings API endpoint along with a choice of embedding model ID (e.g., text-embedding-ada-002). The response will contain an embedding, which you can extract, save, and use.

Public Members

uint32_t (*getUsage)(struct OpenAI_EmbeddingResponse *stringResponse)

Get the usage of OpenAI_EmbeddingResponse.

Param stringResponse[in]: The pointer to OpenAI_EmbeddingResponse
Return: uint32_t

uint32_t (*getLen)(struct OpenAI_EmbeddingResponse *embeddingData)

Get the length of OpenAI_EmbeddingResponse.

Param embeddingData[in]: The pointer to OpenAI_EmbeddingResponse
Return: uint32_t

OpenAI_EmbeddingData_t *(*getData)(struct OpenAI_EmbeddingResponse *embeddingData, uint32_t index)

Get the data of OpenAI_EmbeddingResponse.

Param embeddingData[in]: The pointer to OpenAI_EmbeddingResponse
Param index[in]: The index of the data
Return: OpenAI_EmbeddingData_t*

char *(*getError)(struct OpenAI_EmbeddingResponse *embeddingData)

Get the error of OpenAI_EmbeddingResponse.

Param embeddingData[in]: The pointer to OpenAI_EmbeddingResponse
Return: char*

void (*deleteResponse)(struct OpenAI_EmbeddingResponse *embeddingData)

delete the embedding response, should free it after use.

Param embeddingData[in]: the point of OpenAI_EmbeddingResponse

struct OpenAI_ModerationResponse

The moderations endpoint is a tool you can use to check whether content complies with OpenAI’s usage policies. Developers can thus identify content that our usage policies prohibits and take action, for instance by filtering it.

Public Members

uint32_t (*getLen)(struct OpenAI_ModerationResponse *moderationResponse)

Get the length of OpenAI_ModerationResponse.

Param moderationResponse[in]: The pointer to OpenAI_ModerationResponse
Return: uint32_t

bool (*getData)(struct OpenAI_ModerationResponse *moderationResponse, uint32_t index)

Get the moderation result of OpenAI_ModerationResponse.

Param moderationResponse[in]: The pointer to OpenAI_ModerationResponse
Param index[in]: The index of the moderation result
Return: bool

char *(*getError)(struct OpenAI_ModerationResponse *moderationResponse)

Get the error message of OpenAI_ModerationResponse.

Param moderationResponse[in]: The pointer to OpenAI_ModerationResponse
Return: char*

void (*deleteResponse)(struct OpenAI_ModerationResponse *moderationResponse)

delete the moderation response, should free it after use.

Param moderationResponse[in]: the point of } OpenAI_ModerationResponse_t

struct OpenAI_ImageResponse

Save the image which is generated by OpenAI.

Public Members

uint32_t (*getLen)(struct OpenAI_ImageResponse *imageResponse)

Get the length of OpenAI_ImageResponse.

Param imageResponse[in]: The pointer to OpenAI_ImageResponse
Return: uint32_t

char *(*getData)(struct OpenAI_ImageResponse *imageResponse, uint32_t index)

Get the data of OpenAI_ImageResponse.

Param imageResponse[in]: The pointer to OpenAI_ImageResponse
Param index[in]: The index of the image data
Return: char*

char *(*getError)(struct OpenAI_ImageResponse *imageResponse)

Get the error message of OpenAI_ImageResponse.

Param imageResponse[in]: The pointer to OpenAI_ImageResponse
Return: char*

void (*deleteResponse)(struct OpenAI_ImageResponse *imageResponse)

delete the image response

Param imageResponse[in]: the point of } OpenAI_ImageResponse_t

struct OpenAI_StringResponse

Parse the returned json data into OpenAI_StringResponse_t.

Public Members

uint32_t (*getUsage)(struct OpenAI_StringResponse *stringResponse)

get the usage of openai response

Param stringResponse[in]: the point of OpenAI_StringResponse_t
Return: uint32_t

uint32_t (*getLen)(struct OpenAI_StringResponse *stringResponse)

get the len of openai response

Param stringResponse[in]: the point of OpenAI_StringResponse_t
Return: uint32_t

char *(*getData)(struct OpenAI_StringResponse *stringResponse, uint32_t index)

get the data of openai response

Param stringResponse[in]: the point of OpenAI_StringResponse_t
Param index[in]: the index of data
Return: char*

char *(*getError)(struct OpenAI_StringResponse *stringResponse)

get the error of openai response

Param stringResponse[in]: the point of OpenAI_StringResponse_t
Return: char*

void (*deleteResponse)(struct OpenAI_StringResponse *stringResponse)

delete the openai response

Param stringResponse[in]: the point of OpenAI_StringResponse_t

struct OpenAI_SpeechResponse

Store the returned data into a OpenAI_SpeechResponse_t structure.

Public Members

uint32_t (*getLen)(struct OpenAI_SpeechResponse *SpeechResponse)

get the len of openai speech response

Param speechResponse[in]: the point of OpenAI_SpeechResponse_t
Return: uint32_t

char *(*getData)(struct OpenAI_SpeechResponse *SpeechResponse)

get the data of openai response

Param SpeechResponse[in]: the point of OpenAI_SpeechResponse_t
Return: char*

void (*deleteResponse)(struct OpenAI_SpeechResponse *SpeechResponse)

delete the openai response

Param SpeechResponse[in]: the point of OpenAI_SpeechResponse_t

struct OpenAI_Completion

Given a prompt, the model will return one or more predicted completions, and can also return the probabilities of alternative tokens at each position.

Public Members

void (*setModel)(struct OpenAI_Completion *completion, const char *m)

Set the model to use for completion.

Param completion[in]: the point of OpenAI_Completion_t
Param m[in]: the name of the model to use for completion

void (*setMaxTokens)(struct OpenAI_Completion *completion, uint32_t mt)

Set the maximum number of tokens to generate in the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param mt[in]: the maximum number of tokens to generate in the completion

void (*setTemperature)(struct OpenAI_Completion *completion, float t)

Set the temperature of the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param t[in]: float between 0 and 2. Higher value gives more random results.

void (*setTopP)(struct OpenAI_Completion *completion, float tp)

Set the value of top_p for the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param tp[in]: float between 0 and 1. recommended to alter this or temperature but not both.

void (*setN)(struct OpenAI_Completion *completion, uint32_t n)

Set the number of completions to generate for each prompt.

Param completion[in]: the point of OpenAI_Completion_t
Param n[in]: the number of completions to generate for each prompt

void (*setEcho)(struct OpenAI_Completion *completion, bool e)

Echo back the prompt in addition to the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param e[in]: true if the prompt should be echoed back, false otherwise

void (*setStop)(struct OpenAI_Completion *completion, const char *s)

Set up to 4 sequences where the API will stop generating further tokens.

Param completion[in]: the point of OpenAI_Completion_t
Param s[in]: the sequences where the API will stop generating further tokens

void (*setPresencePenalty)(struct OpenAI_Completion *completion, float pp)

Set the presence penalty for the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param pp[in]: float between -2.0 and 2.0. Positive values increase the model’s likelihood to talk about new topics.

void (*setFrequencyPenalty)(struct OpenAI_Completion *completion, float fp)

Set the frequency penalty for the completion.

Param completion[in]: the point of OpenAI_Completion_t
Param fp[in]: float between -2.0 and 2.0. Positive values decrease the model’s likelihood to repeat the same line verbatim.

void (*setBestOf)(struct OpenAI_Completion *completion, uint32_t bo)

Generates best_of completions server-side and returns the “best”. “best_of” must be greater than “n”.

Param completion[in]: the point of OpenAI_Completion_t
Param bo[in]: the number of best_of completions to generate server-side and return the “best”

void (*setUser)(struct OpenAI_Completion *completion, const char *u)

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Param completion[in]: the point of OpenAI_Completion_t
Param u[in]: the unique identifier representing your end-user

OpenAI_StringResponse_t *(*prompt)(struct OpenAI_Completion *completion, char *p)

Send the prompt for completion.

Param completion[in]: the point of OpenAI_Completion_t
Param p[in]: the prompt for completion
Return: OpenAI_StringResponse_t*

struct OpenAI_ChatCompletion

Given a list of messages comprising a conversation, the model will return a response.

Public Members

void (*setModel)(struct OpenAI_ChatCompletion *chatCompletion, const char *m)

Set the model to use for completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param m[in]: the name of the model to use for chatCompletion

void (*setSystem)(struct OpenAI_ChatCompletion *chatCompletion, const char *s)

Set the system to use for completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param s[in]: description of the required assistant

void (*setMaxTokens)(struct OpenAI_ChatCompletion *chatCompletion, uint32_t mt)

Set the maximum number of tokens to generate in the completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param mt[in]: the maximum number of tokens to generate in the completion

void (*setTemperature)(struct OpenAI_ChatCompletion *chatCompletion, float t)

Set the temperature for the completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param t[in]: float between 0 and 2. Higher value gives more random results.

void (*setTopP)(struct OpenAI_ChatCompletion *chatCompletion, float tp)

Set the top_p for the completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param tp[in]: float between 0 and 1. recommended to alter this or temperature but not both.

void (*setStop)(struct OpenAI_ChatCompletion *chatCompletion, const char *s)

Set up to 4 sequences where the API will stop generating further tokens.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param s[in]: the sequences where the API will stop generating further tokens

void (*setPresencePenalty)(struct OpenAI_ChatCompletion *chatCompletion, float pp)

Set the presence penalty for the completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param pp[in]: float between -2.0 and 2.0. Positive values increase the model’s likelihood to talk about new topics.

void (*setFrequencyPenalty)(struct OpenAI_ChatCompletion *chatCompletion, float fp)

Set the frequency penalty for the completion.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param fp[in]: float between -2.0 and 2.0. Positive values decrease the model’s likelihood to repeat the same line verbatim.

void (*setUser)(struct OpenAI_ChatCompletion *chatCompletion, const char *u)

A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param u[in]: the unique identifier representing your end-user

void (*clearConversation)(struct OpenAI_ChatCompletion *chatCompletion)

Clears the accumulated conversation.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion

OpenAI_StringResponse_t *(*message)(struct OpenAI_ChatCompletion *chatCompletion, const char *p, bool save)

Send the message for completion. Save it with the first response if selected.

Param chatCompletion[in]: the point of OpenAI_ChatCompletion
Param p[in]: the message for completion
Param save[in]: save it with the first response if selected
Return: OpenAI_StringResponse_t*

struct OpenAI_Edit

Given a prompt and an instruction, the model will return an edited version of the prompt.

Public Members

void (*setModel)(struct OpenAI_Edit *edit, const char *m)

Set the model to use for edit.

Param edit[in]: the point of OpenAI_Edit_t
Param m[in]: the name of the model to use for edit

void (*setTemperature)(struct OpenAI_Edit *edit, float t)

Set the temperature for the edit.

Param edit[in]: the point of OpenAI_Edit_t
Param t[in]: float between 0 and 2. Higher value gives more random results.

void (*setTopP)(struct OpenAI_Edit *edit, float tp)

Set the top_p for the edit.

Param edit[in]: the point of OpenAI_Edit_t
Param tp[in]: float between 0 and 1. recommended to alter this or temperature but not both.

void (*setN)(struct OpenAI_Edit *edit, uint32_t n)

Set the number of edits to generate for the input and instruction.

Param edit[in]: the point of OpenAI_Edit_t
Param n[in]: the number of edits to generate for the input and instruction

OpenAI_StringResponse_t *(*process)(struct OpenAI_Edit *edit, char *instruction, char *input)

Creates a new edit for the provided input, instruction, and parameters.

Param edit[in]: the point of OpenAI_Edit_t
Param instruction[in]: the instruction for the edit
Param input[in]: the input text to be edited
Return: OpenAI_StringResponse_t* the edited text

struct OpenAI_ImageGeneration

Creates an image given a prompt.

Public Members

void (*setSize)(struct OpenAI_ImageGeneration *imageGeneration, OpenAI_Image_Size s)

Set the size of the generated images.

Param imageGeneration[in]: the point of OpenAI_ImageGeneration
Param s[in]: the size of the generated images

void (*setResponseFormat)(struct OpenAI_ImageGeneration *imageGeneration, OpenAI_Image_Response_Format rf)

Set the format in which the generated images are returned.

Param imageGeneration[in]: the point of OpenAI_ImageGeneration
Param rf[in]: the format in which the generated images are returned

void (*setN)(struct OpenAI_ImageGeneration *imageGeneration, uint32_t n)

Set the number of images to generate. Must be between 1 and 10.

Param imageGeneration[in]: the point of OpenAI_ImageGeneration
Param n[in]: the number of images to generate

void (*setUser)(struct OpenAI_ImageGeneration *imageGeneration, const char *u)

Set a unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Param imageGeneration[in]: the point of OpenAI_ImageGeneration
Param u[in]: the unique identifier representing your end-user

OpenAI_ImageResponse_t *(*prompt)(struct OpenAI_ImageGeneration *imageGeneration, char *p)

Creates image/images from given a prompt.

Param imageGeneration[in]: the point of OpenAI_ImageGeneration
Param p[in]: the prompt for image generation
Return: OpenAI_ImageResponse_t* the generated image/images

struct OpenAI_ImageVariation

Creates a variation of a given image.

Public Members

void (*setSize)(struct OpenAI_ImageVariation *imageVariation, OpenAI_Image_Size s)

Set the size of the generated images.

Param imageVariation[in]: the point of OpenAI_ImageVariation
Param s[in]: the size of the generated images

void (*setResponseFormat)(struct OpenAI_ImageVariation *imageVariation, OpenAI_Image_Response_Format rf)

Set the format in which the generated images are returned.

Param imageVariation[in]: the point of OpenAI_ImageVariation
Param rf[in]: the format in which the generated images are returned

void (*setN)(struct OpenAI_ImageVariation *imageVariation, uint32_t n)

Set the number of images to generate. Must be between 1 and 10.

Param imageVariation[in]: the point of OpenAI_ImageVariation
Param n[in]: the number of images to generate

void (*setUser)(struct OpenAI_ImageVariation *imageVariation, const char *u)

Set a unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Param imageVariation[in]: the point of OpenAI_ImageVariation
Param u[in]: the unique identifier representing your end-user

OpenAI_ImageResponse_t *(*image)(struct OpenAI_ImageVariation *imageVariation, uint8_t *data, size_t len)

Creates an image variation from given image data.

Param imageVariation[in]: the point of OpenAI_ImageVariation
Param data[in]: the input image data
Param len[in]: the length of the input image data
Return: OpenAI_ImageResponse_t* the generated image variation

struct OpenAI_ImageEdit

Creates an edited or extended image given an original image and a prompt.

Public Members

void (*setPrompt)(struct OpenAI_ImageEdit *imageEdit, const char *p)

Set the prompt for the image edit.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param p[in]: the prompt for the image edit

void (*setSize)(struct OpenAI_ImageEdit *imageEdit, OpenAI_Image_Size s)

Set the size of the generated images.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param s[in]: the size of the generated images

void (*setResponseFormat)(struct OpenAI_ImageEdit *imageEdit, OpenAI_Image_Response_Format rf)

Set the format in which the generated images are returned.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param rf[in]: the format in which the generated images are returned

void (*setN)(struct OpenAI_ImageEdit *imageEdit, uint32_t n)

Set the number of images to generate. Must be between 1 and 10.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param n[in]: the number of images to generate

void (*setUser)(struct OpenAI_ImageEdit *imageEdit, const char *u)

Set a unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param u[in]: the unique identifier representing your end-user

OpenAI_ImageResponse_t *(*image)(struct OpenAI_ImageEdit *imageEdit, uint8_t *data, size_t len, uint8_t *mask_data, size_t mask_len)

Creates an edited or extended image given an original image, a mask, and a prompt.

Param imageEdit[in]: the point of OpenAI_ImageEdit_t
Param data[in]: the input image data
Param len[in]: the length of the input image data
Param mask_data[in]: the input mask data
Param mask_len[in]: the length of the input mask data
Return: OpenAI_ImageResponse_t* the edited or extended image

struct OpenAI_AudioTranscription

Transcribes audio into the input language.

Public Members

void (*setPrompt)(struct OpenAI_AudioTranscription *audioTranscription, const char *p)

Set the prompt for the audio transcription.

Param audioTranscription[in]: the point of OpenAI_AudioTranscription_t
Param p[in]: the prompt for the audio transcription

void (*setResponseFormat)(struct OpenAI_AudioTranscription *audioTranscription, OpenAI_Audio_Response_Format rf)

Set the format of the transcript output.

Param audioTranscription[in]: the point of OpenAI_AudioTranscription_t
Param rf[in]: the format of the transcript output

void (*setTemperature)(struct OpenAI_AudioTranscription *audioTranscription, float t)

Set the temperature for the audio transcription.

Param audioTranscription[in]: the point of OpenAI_AudioTranscription_t
Param t[in]: float between 0 and 1

void (*setLanguage)(struct OpenAI_AudioTranscription *audioTranscription, const char *l)

Set the language of the input audio.

Param audioTranscription[in]: the point of OpenAI_AudioTranscription_t
Param l[in]: the language in ISO-639-1 format of the input audio. NULL for Auto.

char *(*file)(struct OpenAI_AudioTranscription *audioTranscription, uint8_t *data, size_t len, OpenAI_Audio_Input_Format f)

Transcribe an audio file.

Param audioTranscription[in]: the point of OpenAI_AudioTranscription_t
Param data[in]: the input audio data
Param len[in]: the length of the input audio data
Param f[in]: the format of the input audio data
Return: char* the transcribed text, you should free it after use.

struct OpenAI_AudioSpeech

Given a list of messages comprising a conversation, the model will return a response.

Public Members

void (*setModel)(struct OpenAI_AudioSpeech *createSpeech, const char *m)

Set the model to use for completion.

Param createSpeech[in]: the point of OpenAI_SpeechResponse_t
Param m[in]: the name of the model to use for audio response

void (*setVoice)(struct OpenAI_AudioSpeech *createSpeech, const char *m)

Set the voice to use for completion.

Param createSpeech[in]: the point of OpenAI_SpeechResponse_t
Param m[in]: the name of the model to use for audio response

void (*setSpeed)(struct OpenAI_AudioSpeech *createSpeech, float t)

Set the speed of the output audio.

Param createSpeech[in]: the point of OpenAI_SpeechResponse_t
Param t[in]: float between 0.25 to 4.0

void (*setResponseFormat)(struct OpenAI_AudioSpeech *createSpeech, OpenAI_Audio_Output_Format rf)

Set the format of the output.

Param createSpeech[in]: the point of OpenAI_SpeechResponse_t
Param rf[in]: the format of the output audio

OpenAI_SpeechResponse_t *(*speech)(struct OpenAI_AudioSpeech *createSpeech, char *p)

Send the message for completion. Save it with the first response if selected.

Param createSpeech[in]: the point of OpenAI_SpeechResponse_t
Param p[in]: the message for audio generation
Return: *

struct OpenAI_AudioTranslation

Translates audio into English.

Public Members

void (*setPrompt)(struct OpenAI_AudioTranslation *audioTranslation, const char *p)

Set the prompt for the audio translation.

Param audioTranslation[in]: the point of OpenAI_AudioTranslation_t
Param p[in]: the prompt for the audio translation

void (*setResponseFormat)(struct OpenAI_AudioTranslation *audioTranslation, OpenAI_Audio_Response_Format rf)

Set the format of the transcript output.

Param audioTranslation[in]: the point of OpenAI_AudioTranslation_t
Param rf[in]: the format of the transcript output

void (*setTemperature)(struct OpenAI_AudioTranslation *audioTranslation, float t)

Set the temperature for the audio translation.

Param audioTranslation[in]: the point of OpenAI_AudioTranslation_t
Param t[in]: float between 0 and 2. Higher value gives more random results.

char *(*file)(struct OpenAI_AudioTranslation *audioTranslation, uint8_t *data, size_t len, OpenAI_Audio_Input_Format f)

Transcribe and translate an audio file into English.

Param audioTranslation[in]: the point of OpenAI_AudioTranslation_t
Param data[in]: the input audio data
Param len[in]: the length of the input audio data
Param f[in]: the format of the input audio data
Return: char* the translated text in English, you should free it after use.

struct OpenAI: The entry point for calling the Openai api.

Type Definitions

typedef struct OpenAI_EmbeddingResponse OpenAI_EmbeddingResponse_t: To get an embedding, send your text string to the embeddings API endpoint along with a choice of embedding model ID (e.g., text-embedding-ada-002). The response will contain an embedding, which you can extract, save, and use.

typedef struct OpenAI_ModerationResponse OpenAI_ModerationResponse_t: The moderations endpoint is a tool you can use to check whether content complies with OpenAI’s usage policies. Developers can thus identify content that our usage policies prohibits and take action, for instance by filtering it.

typedef struct OpenAI_ImageResponse OpenAI_ImageResponse_t: Save the image which is generated by OpenAI.

typedef struct OpenAI_StringResponse OpenAI_StringResponse_t: Parse the returned json data into OpenAI_StringResponse_t.

typedef struct OpenAI_SpeechResponse OpenAI_SpeechResponse_t: Store the returned data into a OpenAI_SpeechResponse_t structure.

typedef struct OpenAI_Completion OpenAI_Completion_t: Given a prompt, the model will return one or more predicted completions, and can also return the probabilities of alternative tokens at each position.

typedef struct OpenAI_ChatCompletion OpenAI_ChatCompletion_t: Given a list of messages comprising a conversation, the model will return a response.

typedef struct OpenAI_Edit OpenAI_Edit_t: Given a prompt and an instruction, the model will return an edited version of the prompt.

typedef struct OpenAI_ImageGeneration OpenAI_ImageGeneration_t: Creates an image given a prompt.

typedef struct OpenAI_ImageVariation OpenAI_ImageVariation_t: Creates a variation of a given image.

typedef struct OpenAI_ImageEdit OpenAI_ImageEdit_t: Creates an edited or extended image given an original image and a prompt.

typedef struct OpenAI_AudioTranscription OpenAI_AudioTranscription_t: Transcribes audio into the input language.

typedef struct OpenAI_AudioSpeech OpenAI_AudioSpeech_t: Given a list of messages comprising a conversation, the model will return a response.

typedef struct OpenAI_AudioTranslation OpenAI_AudioTranslation_t: Translates audio into English.

typedef struct OpenAI OpenAI_t: The entry point for calling the Openai api.

Enumerations

enum OpenAI_Image_Size

Values:

enumerator OPENAI_IMAGE_SIZE_1024x1024

enumerator OPENAI_IMAGE_SIZE_512x512

enumerator OPENAI_IMAGE_SIZE_256x256

enum OpenAI_Image_Response_Format

Values:

enumerator OPENAI_IMAGE_RESPONSE_FORMAT_URL

enumerator OPENAI_IMAGE_RESPONSE_FORMAT_B64_JSON

enum OpenAI_Audio_Response_Format

Values:

enumerator OPENAI_AUDIO_RESPONSE_FORMAT_JSON

enumerator OPENAI_AUDIO_RESPONSE_FORMAT_TEXT

enumerator OPENAI_AUDIO_RESPONSE_FORMAT_SRT

enumerator OPENAI_AUDIO_RESPONSE_FORMAT_VERBOSE_JSON

enumerator OPENAI_AUDIO_RESPONSE_FORMAT_VTT

enum OpenAI_Audio_Input_Format

Values:

enumerator OPENAI_AUDIO_INPUT_FORMAT_MP3

enumerator OPENAI_AUDIO_INPUT_FORMAT_MP4

enumerator OPENAI_AUDIO_INPUT_FORMAT_MPEG

enumerator OPENAI_AUDIO_INPUT_FORMAT_MPGA

enumerator OPENAI_AUDIO_INPUT_FORMAT_M4A

enumerator OPENAI_AUDIO_INPUT_FORMAT_WAV

enumerator OPENAI_AUDIO_INPUT_FORMAT_WEBM

enum OpenAI_Audio_Output_Format

Values:

enumerator OPENAI_AUDIO_OUTPUT_FORMAT_MP3

enumerator OPENAI_AUDIO_OUTPUT_FORMAT_OPUS

enumerator OPENAI_AUDIO_OUTPUT_FORMAT_AAC

enumerator OPENAI_AUDIO_OUTPUT_FORMAT_FLAC

提供有关此文档的反馈