relic-yuexi / jinaai-py

JinaAI Python SDK

The JinaAI Python SDK is an efficient instrument that smoothly brings the power of JinaAI's products — SceneXplain, PromptPerfect, Rationale, BestBanner, and JinaChat — into Python applications. Acting as a sturdy interface for JinaAI's APIs, this SDK lets you effortlessly formulate and fine-tune prompts, thus streamlining application development.

Installing

Package manager

Using pip:

$ pip install jinaai

API secrets

To generate an API secret, you need to authenticate on each respective platform's API tab:

• SceneXplain API
• PromptPerfect API
• Rationale API
• JinaChat API
• BestBanner API

Note: Each secret is product-specific and cannot be interchanged. If you're planning to use multiple products, you'll need to generate a separate secret for each.

Example usage

Import the SDK and instantiate a new client with your authentication secrets:

from jinaai import JinaAI

jinaai = JinaAI(
    secrets = {
        'promptperfect-secret': 'XXXXXX',
        'scenex-secret': 'XXXXXX',
        'rationale-secret': 'XXXXXX',
        'jinachat-secret': 'XXXXXX',
        'bestbanner-secret': 'XXXXXX',
    }
)

Describe images:

descriptions = jinaai.describe(
    'https://picsum.photos/200'
)

Evaluate situations:

decisions = jinaai.decide(
    'Going to Paris this summer', 
    { 'analysis': 'proscons' }
)

Optimize prompts:

prompts = jinaai.optimize(
    'Write an Hello World function in Python'
)

Generate complex answers:

output = jinaai.generate(
    'Give me a recipe for a pizza with pineapple'
)

Create images from text:

output = jinaai.imagine(
    'A controversial fusion of sweet pineapple and savory pizza.'
)

Use APIs together:

situations = [toBase64(img) for img in [
    'factory-1.png',
    'factory-2.png',
    'factory-3.png',
    'factory-4.png',
]]

descriptions = jinaai.describe(situations)

prompt1 = [
    'Do any of those situations present a danger?',
    'Reply with [YES] or [NO] and explain why',
    *['SITUATION:\n' + desc['output'] for i, desc in enumerate(descriptions['results'])]
]

analysis = jinaai.generate('\n'.join(prompt1))

prompt2 = [
    'What should be done first to make those situations safer?',
    'I only want the most urgent situation',
    *['SITUATION:\n' + desc['output'] for i, desc in enumerate(descriptions['results'])]
]

recommendation = jinaai.generate('\n'.join(propmt2))

swot = jinaai.decide(
    recommendation['output'],
    { 'analysis': 'swot' }
)

banners = jinaai.imagine(
    *[desc['output'] for i, desc in enumerate(descriptions['results'])]
)

Raw Output

You can retrieve the raw output of each APIs by passing raw: True in the options:

descriptions = jinaai.describe(
    'https://picsum.photos/200',
    { 'raw': True }
)

print(descriptions['raw'])

Custom Base Urls

Custom base Urls can be passed directly in the client's constructor:

jinaai = JinaAI(
    baseUrls={
        'promptperfect': 'https://promptperfect-customurl.jina.ai',
        'scenex': 'https://scenex-customurl.jina.ai',
        'rationale': 'https://rationale-customurl.jina.ai',
        'jinachat': 'https://jinachat-customurl.jina.ai',
        'bestbanner': 'https://bestbanner-customurl.jina.ai',
    }
)

API Documentation

JinaAi.describe

output = JinaAI.describe(input, options)

• Input

VARIABLE	TYPE	VALUE
input	str / str array	Image URL or Base64

• Options

VARIABLE	TYPE	VALUE
options	dict
options['algorithm']	None / str	Aqua / Bolt / Comet / Dune / Ember / Flash / Glide / Hearth / Inception / Jelly
options['features']	None / str array	high_quality, question_answer, tts, opt-out, json
options['languages']	None / str array	en, cn, de, fr, it...
options['question']	None / str	Question related to the picture(s)
options['style']	None / str	default / concise / prompt
options['output_length']	None / number
options['json_schema']	None / dict
options['callback_url']	None / string

• Output

VARIABLE	TYPE	VALUE
output	dict
output['results']	dict array
results[0]['output']	str	The picture description
results[0]['i18n']	dict	Contains one key for each item in languages
...i18n['cn']	str	The translated picture description
...i18n['cn']	dict array	Only for Hearth algorithm
...i18n['cn'][0]	dict
...i18n['cn'][0]['message']	str
...i18n['cn'][0]['isNarrator']	boolean
...i18n['cn'][0]['name']	str
...i18n['cn']	dict array	Only for Inception algorithm
...i18n['cn'][0]	dict
...i18n['cn'][0]['summary']	str
...i18n['cn'][0]['events']	dict array
...['events']['description']	str
...['events']['timestamp']	str
results[0]['tts']	dict	Only for Hearth algorithm
...tts['cn']	str	Contains the url to the tts file
results[0]['ssml']	dict	Only for Hearth algorithm
...ssml['cn']	str	Contains the url to the ssml file

JinaAi.optimize

output = JinaAI.optimize(input, options)

• Input

VARIABLE	TYPE	VALUE
input	str / str array	Image URL or Base64 / prompt to optimize

• Options

VARIABLE	TYPE	VALUE
options	dict
options['targetModel']	None / str	chatgpt / gpt-4 / stablelm-tuned-alpha-7b / claude / cogenerate / text-davinci-003 / dalle / sd / midjourney / kandinsky / lexica
options['features']	None / str array	preview, no_spam, shorten, bypass_ethics, same_language, always_en, high_quality, redo_original_image, variable_subs, template_run
options['iterations']	None / number	Default: 1
options['previewSettings']	None / dict	Contains the settings for the preview
...previewSettings['temperature']	number	Example: 0.9
...previewSettings['topP']	number	Example: 0.9
...previewSettings['topK']	number	Example: 0
...previewSettings['frequencyPenalty']	number	Example: 0
...previewSettings['presencePenalty']	number	Example: 0
options['previewVariables']	None / dict	Contains one key for each variables in the prompt
...previewVariables['var1']	str	The value of the variable
options['timeout']	Number	Default: 20000
options['target_language']	None / str	en / cn / de / fr / it...

• Output

VARIABLE	TYPE	VALUE
output	dict
output['results']	dict array
results[0]['output']	str	The optimized prompt

JinaAi.decide

output = JinaAI.decide(input, options)

• Input

VARIABLE	TYPE	VALUE
input	str / str array	Decision to evaluate

• Options

VARIABLE	TYPE	VALUE
options	dict
options['analysis']	None / str	proscons / swot / multichoice / outcomes
options['style']	None / str	concise / professional / humor / sarcastic / childish / genZ
options['profileId']	None / str	The id of the Personas you want to use

• Output

VARIABLE	TYPE	VALUE
output	dict
output['results']	dict array
results[0]['proscons']	None / dict
...proscons['pros']	dict	Contains one key for each pros
...proscons['pros']['pros1']	str	The explanation of the pros
...proscons['cons']	dict	Contains one key for each cons
...proscons['cons']['cons1']	str	The explanation of the cons
...proscons['bestChoice']	str
...proscons['conclusion']	str
...proscons['confidenceScore']	number
results[0]['swot']	None / dict
...swot['strengths']	dict	Contains one key for each strength
...swot['strengths']['str1']	str	The explanation of the strength
...swot['weaknesses']	dict	Contains one key for each weakness
...swot['weaknesses']['weak1']	str	The explanation of the weakness
...swot['opportunities']	dict	Contains one key for each opportunity
...swot['opportunities']['opp1']	str	The explanation of the opportunity
...swot['threats']	dict	Contains one key for each threat
...swot['threats']['thre1']	str	The explanation of the threat
...swot['bestChoice']	str
...swot['conclusion']	str
...swot['confidenceScore']	number
results[0]['multichoice']	None / dict	Contains one key for each choice
...multichoice['choice1']	str	The value of the choice
results[0]['outcomes']	None / dict array
...outcomes[0]['children']	None / dict array	a recursive array of results['outcomes']
...outcomes[0]['label']	str
...outcomes[0]['sentiment']	str

JinaAi.generate

output = JinaAI.generate(input, options)

• Input

VARIABLE	TYPE	VALUE
input	str / str array	Image URL or Base64 / prompt

• Options

VARIABLE	TYPE	VALUE
options	dict
options['role']	None / str	user / assistant
options['name']	None / str	The name of the author of this message
options['chatId']	None / str	The id of the conversation to continue
options['stream']	None / boolean	Whether to stream back partial progress, Default: false
options['temperature']	None / number	Default: 1
options['top_p']	None / str	Default: 1
options['stop']	None / str / str array	Up to 4 sequences where the API will stop generating further tokens
options['max_tokens']	None / number	Default: infinite
options['presence_penalty']	None / number	Number between -2.0 and 2.0, Default: 0
options['frequency_penalty']	None / number	Number between -2.0 and 2.0, Default: 0
options['logit_bias']	None / dict	The likelihood for a token to appear in the completion
...logit_bias['tokenId']	number	Bias value from -100 to 100
options['image']	str	The attached image of the message. The image can be either a URL or a base64-encoded string

• Output

VARIABLE	TYPE	VALUE
output	dict
output['output']	str	The generated answer
output['chatId']	str	The chatId to continue the conversation

JinaAi.imagine

output = JinaAI.imagine(input, options)

• Input

VARIABLE	TYPE	VALUE
input	str / str array	Prompt

• Options

VARIABLE	TYPE	VALUE
options	dict
options['style']	None / str	default / photographic / minimalist / flat

• Output

VARIABLE	TYPE	VALUE
output	dict
output['results']	dict array
results[0]['output']	array	array of 4 image urls

JinaAi.utils

outout = JinaAI.utils.image_to_base64(input)

VARIABLE	TYPE	VALUE
input	str	Image path on disk
output	str	Base64 image

outout = JinaAI.utils.is_url(input)

VARIABLE	TYPE	VALUE
input	str
output	boolean

outout = JinaAI.utils.is_base64(input)

VARIABLE	TYPE	VALUE
input	str
output	boolean