# Suppress extra warning messages to keep things clean
import warnings; warnings.filterwarnings("ignore")
import numpy as np
np.random.seed(42)
 

# Install the OpenAI package so we can use the API
!pip install --quiet openai
 

# Always import the package before using it
from openai import OpenAI
import os
 

# Get the API key from your environment
API_key = os.environ["OPENAI_API_KEY"]
 

# Make the OpenAI client instance for talking to the API
client = OpenAI(api_key=API_key)
 

# Basic Chat Completion  Hello World
response = client.chat.completions.create(
    model="gpt-4.1-mini",
    messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)
 

# Try changing the prompt
prompt = input("What do you want to ask the chat model? ")
custom_response = client.chat.completions.create(
    model="gpt-4.1-mini",
    messages=[{"role": "user", "content": prompt}]
)
print(custom_response.choices[0].message.content)
 

# Let's try the responses endpoint for single answer tasks
resp = client.responses.create(
    model="gpt-4.1-mini",
    input="Summarize what an API key is."
)
print(resp.output[0].content[0].text)
 

# What other models are available?
available_models = [
    "gpt-4o",
    "gpt-4.1",
    "gpt-4.1-mini",
    "text-embedding-3-small",
    "whisper-1",
    "gpt-4o-mini-tts",
]
print("Some OpenAI models:", available_models)
 

# Get vector embeddings for search and machine learning
embed = client.embeddings.create(
    model="text-embedding-3-small",
    input=["Machine learning is fun"]
)
print(len(embed.data[0].embedding))
 

# Speech to text with Whisper
audio_file = open('sample.wav', 'rb')
transcript = client.audio.transcriptions.create(
    model="whisper-1",
    file=audio_file
)
print(transcript.text)
 

# Turn text into speech with TTS
with open('voice.mp3', 'wb') as f:
    audio = client.audio.speech.create(
        model='gpt-4o-mini-tts',
        voice='alloy',
        input='Hello world!'
    )
    f.write(audio.read())
 

# Example: Catching an API error
try:
    broken_response = client.chat.completions.create(
        model="bad-model-name",
        messages=[{"role": "user", "content": "hi"}]
    )
except Exception as error:
    print("Something went wrong:", error)
 

# API Rate Limits  Slow down if you send too many requests
import time
for i in range(3):
    try:
        short_resp = client.responses.create(
            model="gpt-4.1-mini",
            input="Say round number {}.".format(i))
        print(short_resp.output[0].content[0].text)
    except Exception as e:
        print("Please wait. Error:", e)
        time.sleep(2)
 

# Best Practice: Always check your replies
out = client.responses.create(
    model="gpt-4.1-mini",
    input="What is the capital of France?"
)
answer = out.output[0].content[0].text
if answer and "Paris" in answer:
    print("Looks correct:", answer)
else:
    print("Unexpected answer:", answer)
 

# Mini Project: Make a Function Calling Chatbot
def add(a, b):
    return a + b

resp = client.chat.completions.create(
    model="gpt-4.1",
    messages=[{"role": "user", "content": "Add 3 and 5"}],
    functions=[
        {
            "name": "add",
            "description": "Add two numbers",
            "parameters": {
                "type": "object",
                "properties": {
                    "a": {"type": "number"},
                    "b": {"type": "number"}
                },
                "required": ["a", "b"]
            }
        }
    ],
    function_call="auto"
)
print(resp)
 

# Challenge: Ask the user for their own math question
user_math = input("Give me a math question to solve, like 'Add 10 and 15': ")
math_response = client.chat.completions.create(
    model="gpt-4.1",
    messages=[{"role": "user", "content": user_math}],
    functions=[
        {
            "name": "add",
            "description": "Add two numbers",
            "parameters": {
                "type": "object",
                "properties": {
                    "a": {"type": "number"},
                    "b": {"type": "number"}
                },
                "required": ["a", "b"]
            }
        }
    ],
    function_call="auto"
)
print(math_response)
 

# Challenge: Make OpenAI tell a joke
joke_resp = client.responses.create(
    model="gpt-4.1-mini",
    input="Tell me a funny programming joke."
)
print(joke_resp.output[0].content[0].text)