Lesson 807 of 1570
AI model families: GPT-5 and what's new
Understand what makes GPT-5 different from GPT-4 and earlier OpenAI models.
Lesson map
What this lesson covers
Learning path
The main moves in order
- 1The big idea
- 2AI and GPT-5: OpenAI's Latest Big Brain
- 3The big idea
- 4AI and GPT-5 Thinking: When OpenAI Pauses to Reason
Concept cluster
Terms to connect while reading
Section 1
The big idea
GPT-5 is OpenAI's flagship model. It's smarter at reasoning, handles longer documents, and is better at code. Knowing what version you're using matters because answers differ.
Some examples
- Use GPT-5 for hard reasoning tasks
- Notice GPT-5's better code than GPT-4
- Compare GPT-5 vs GPT-4 on the same prompt
- Check which model your tool actually uses
Try it!
Find a hard math or logic question. Ask it to GPT-4 and GPT-5 (if you have access). Note where the answers differ. Even if both right, the explanations vary.
Section 2
AI and GPT-5: OpenAI's Latest Big Brain
Section 3
The big idea
GPT-5 is OpenAI's frontier model. It can read images, write code, do math, and reason through hard problems. ChatGPT Plus uses GPT-5 by default for paying users.
Some examples
- Upload a photo and ask GPT-5 to identify what's in it.
- Use GPT-5 for coding when Sonnet gets stuck.
- GPT-5 is great at structured outputs (JSON, tables).
- It's slower than smaller GPTs but more accurate on hard stuff.
Try it!
Try the same hard question on GPT-5 and a smaller model (Haiku or 4o-mini). Compare which one nails it.
Section 4
AI and GPT-5 Thinking: When OpenAI Pauses to Reason
Section 5
The big idea
GPT-5 Thinking pauses to reason step by step before answering. It's slower but way better at math, code, and tricky logic. Use it when the answer matters more than the speed.
Some examples
- Use Thinking for AP Calc problems — it shows its work.
- Use standard GPT-5 for chat, drafting, and quick answers.
- Thinking can take 30+ seconds — chill and let it cook.
- Cost more per token — don't burn it on 'what's the weather'.
Try it!
Pick a tough math problem. Run it on standard GPT-5 and Thinking. Compare answers AND the work shown.
Section 6
GPT-5 Mini vs Flagship: When Smaller Wins
Section 7
The big idea
most prompts do not need the flagship; benchmark on your real data
Some examples
- Use mini for chat support classifying
- Reserve flagship for novel research help
- Run both on the same 50 examples to decide
Try it!
Open your favorite AI tool and try one of the examples above. Pick the one that matches what you are actually working on this week. Spend 10 minutes, no more. Notice what worked and what did not — that's the real lesson.
End-of-lesson quiz
Check what stuck
15 questions · Score saves to your progress.
Tutor
Curious about “AI model families: GPT-5 and what's new”?
Ask anything about this lesson. I’ll answer using just what you’re reading — short, friendly, grounded.
Progress saved locally in this browser. Sign in to sync across devices.
Related lessons
Keep going
Builders · 40 min
Google's Gemini: When It Beats ChatGPT or Claude
Gemini is Google's chatbot. It has some specific strengths that matter for school work.
Builders · 40 min
AI model families: DeepSeek and the China AI scene
Understand DeepSeek and why China's AI models surprised the world.
Builders · 27 min
AI model families: reasoning models (o1, o3, R1)
Understand what 'reasoning models' do differently and when to use them.
