Ataraxia (Blackbox)

Ataraxia is unofficial Python wrapper for Blackbox AI.

Installation

pip install ataraxia

Usage

Chat

from ataraxia import Blackbox, Model

# Use Models.BLACKBOXAI by default
blackbox = Blackbox()
response = blackbox.chat('What do you think about gen alpha?') # ChatResponse<CHAT_ID>
print(response)

# Chat with image
response = blackbox.chat('What do you think about this image?', 'https://example.com/image.png') # from image link
response = blackbox.chat('What do you think about this image?', 'QWt1IFBIUA==') # from base64 string
response = blackbox.chat('What do you think about this image?', 'data:image/png;base64,QWt1IFBIUA==') # from base64 image data
response = blackbox.chat('What do you think about this image?', b'\x89PNG\r\n\x1a\n\x00\x00\x00\rIHDR\x00\x00\x03\xe8\x00\x00') # from image bytes

# Use other model from Model
blackbox = Blackbox(Model.GPT_4O)

# Set system instruction prompt
blackbox = Blackbox(instruction='You are a gen alpha assistant')
# or
blackbox.change_instruction('You are a gen alpha assistant')

# Load chat data and history from saved file
blackbox = Blackbox(chat_id='SAVED_CHAT_ID')
# or
blackbox.load_chat('SAVED_CHAT_ID')

# Save chat data and history to JSON file
blackbox.save_chat()
# or
blackbox.save_chat('./ataraxia_{CHAT_ID}.json') # {CHAT_ID} will be replaced with current chat id

# Change save path
blackbox.change_save_path('filename_{CHAT_ID}.json')

Image Generation (Imagine)

from ataraxia import Blackbox

blackbox = Blackbox()
result = blackbox.imagine('A pregnant kitten') # ImageResponse<IMAGE_ID>
print(result) # https://storage.googleapis.com/...

# Download the image to bytes
image_bytes = result.download()

# Save the image to file
result.save('./imagine.jpg') # True if success False otherwise

Search

from ataraxia import Blackbox

blackbox = Blackbox()
results = blackbox.search('Schall meaning') # list of SearchResult<TITLE>
print(results[0]) # Schall. [ʃal]Maskulinum | masculine m...

Deep Search

from ataraxia import Blackbox

blackbox = Blackbox()
result = blackbox.deep_search('Schall meaning') # SearchResult<OK>
print(result)

Params

Blackbox

Param	Type	Description	Required	Default
model	Models	Set the Blackbox AI (chat) default model	No	Model.BLACKBOXAI
chat_id	str	Set the saved chat_id to load the history and chat data	No	None
instruction	str	Set the system instruction for every chat request	No	None
auto_save	bool	Will save chat data and history to save_path if True	No	True
save_path	str	File path to save chat data like model, instruction, and history. It is recommended to always insert {CHAT_ID} so that it can be loaded with load_chat()	No	./ataraxia_{CHAT_ID}.json

chat

Param	Type	Description	Required	Default
message	str	Message to send	Yes	-
image	str | bytes	Additional image	No	None
history	list	Chat history	No	[]
max_tokens	int	Response max tokens	No	1024
top_p	float	Probability sampling to determine the selected set of tokens for the response	No	0.9
temperature	float	Set the level of creativity or randomness in the generated response	No	0.7

Buy me tanghulu

https://trakteer.id/alfi-syahri