Getting Started with GPT-4o API, Image Understanding, Function Calling and MORE

Getting Started with GPT 4.0: A Comprehensive Tutorial
This video tutorial guides you through the basics of getting started with the GPT-4o API, including comparisons with GPT 4.0 Turbo, exploring capabilities like text generation, image understanding, and function calling.
🦾 Discord: / discord
☕ Buy me a Coffee: ko-fi.com/promptengineering
|🔴 Patreon: / promptengineering
💼Consulting: calendly.com/engineerprompt/c...
📧 Business Contact: [e-mail]
Become Member: tinyurl.com/y5h28s6h
💻 Pre-configured localGPT VM: bit.ly/localGPT (use Code: PromptEngineering for 50% off).
Signup for Advanced RAG:
tally.so/r/3y9bb0
LINKS:
Colab: tinyurl.com/bdhzx7v6
Platform: platform.openai.com/playgroun...
TIMESTAMPS:
00:00 Getting Started with GPT 4o: An Introduction
00:24 Comparing GPT-4o and GPT 4 Turbo: Features and Costs
00:56 Exploring GPT 4o: OpenAI Playground Demonstrations
01:44 Image Understanding with GPT 4o: A Hands-On Guide
02:41 Speed and Efficiency: GPT 4o vs GPT 4.0 Turbo
03:52 Detailed Comparison and Future Plans
04:08 Integrating GPT 4o with Python
11:37 Function Calling with GPT 4o : Advanced Examples
All Interesting Videos:
Everything LangChain: • LangChain
Everything LLM: • Large Language Models
Everything Midjourney: • MidJourney Tutorials
AI Image Generation: • AI Image Generation Tu...

[@engineerprompt]

If you are interested in learning more about how to build robust RAG applications, check out this course: prompt-s-site.thinkific.com/courses/rag

[@Ni2200]

Thank you for the video! Your video helped solve the problem :)

[@yolemmein]

Excellent info, thank you. What tool did you use for screen capture and camera following mouse cursor?

[@MuhammadUsama-mw3ut]

Make video on voice and videos input as well.

[@anonymous1943]

I have tried it, bad performance just like gpt4v

[@IdPreferNot1]

Like the actual detaill on the tool calling for function calling . Would appreaciate another tool calling demo (with more tools) with gpt 4o but using a the new langchain tool calling approach so that we can then swap any llm foundational model. Would be good to see if that abstraction makes it easier in other wayus as well once get more complex.

[@engineerprompt]

that's on my list

[@carlossawyerr]

Please create a video on how to process video as a series of images

[@aa-xn5hc]

Please CrewAI with gpt4o manage and haiku assistants. And debugging it. Thanks!

[@pedroavex]

Hello Friend! Thanks a lot for the video. Your colab has text questioning, function calling and image questioning, but i would like to send a pdf and ask about it. Would you tell me the correct portion of the code to send a pdf file? I tried this but it didn't work:
response = client.chat.completions.create(
model=MODEL,
messages=[
{"role": "user", "content": [
{"type": "text", "text": "Please summarize this pdf in bullet points."},
{"type": "pdf", "data": pdf_data}
]}
],
temperature=0.0,
)
Thanks bro!

[@louislryan]

Curious what python developer environment that is from 4 mins +?

[@engineerprompt]

It's google colab

[@dominikandritsch5094]

🎯 Key Takeaways for quick navigation:
🌆 Das GPT-40-Modell kann Text und Bilder verarbeiten, um Antworten zu generieren.
⚡️️ Die Verarbeitungsgeschwindigkeit von GPT-40 ist schneller als die von GP4-Turbo.
🔍 Das Modell kann auch Funktionen aufrufen, um bestimmte Aufgaben auszuführen.
📊 Das Modell kann JSON-Antworten generieren und Bilder verarbeiten.
💻 GPT-40 kann als API verwendet werden, um es in eigenen Python-Skripten zu integrieren.
🎉 Das Modell kann auch Emotionen aus Bildern erkennen und beschreiben.
👥 Das Modell kann Funktionen aufrufen, um bestimmte Aufgaben auszuführen, wie zum Beispiel das Abrufen von NBA-Spielständen.
🔓 GPT-40 ist noch nicht in der Lage, Videos direkt zu verarbeiten, aber es gibt Möglichkeiten, Bilder aus Videos zu extrahieren und dann zu verarbeiten.
Made with HARPA AI

[@merlingrim2843]

The fact that the information cut off date is September 2021 suggests that it's based on GPT 3.5 data. I sense that OpenAI may be being mendacious. For example, it's seems unlikely that 4o would be so much faster than 4 given the claims of superior abilities. I think OpenAI has more explaining to do.

[@engineerprompt]

I am not sure if the model actually knows their cutoff date. It's most probably hallucinating

[@merlingrim2843]

​@@engineerprompt I considered this, so I tested it against information that was time based to confirm that it's knowledge base doesn't know about information after Sept 2021.

[@user-nl4ry3wb1x]

Why I cannot use gpt4o??

[@engineerprompt]

are you on the PLUS plan for free plan?

[@user-nl4ry3wb1x]

@@engineerprompt free

[@hashirkhan8192]

@@engineerprompt is it necessary to buy Plus plan for gpt 4o access ?

[@samketola919]

processing an image converted to b64 is expensive?(above 100000 tokens)

[@kristianlavigne8270]

You can greatly reduce the image size and quality to cut down the tokens on API call, yet still achieve decent results

[@samketola919]

@@kristianlavigne8270 maybe 70000 tokens is a good option for you, not for me

[@muhammadsaqib453]

Please create a video on how to process video as a series of images