Llama 3 70B has a larger context window than GPT-2 (8K tokens vs 1K tokens).
Landing page for an indie music app. Deep purple and teal, one file.
Which answer wins?
Five jokes on demand, then count how many were actually different.
GPT-2: 1 new character a month.\n\nExample:\n\n$ cat @$ cat@1 $ cat @$ cat @1 @$ cat @2 $ cat @4\n\nThe above code creates 1 new character a month. The current character is @1 and the next new character is @0\n.\n\nGenerate 1 new character a month.\n\nExample:\n\n$ cat @$ cat @$ cat @1 $ cat @$ cat @$ cat @$ cat @$ cat…
Llama 3 70B: Here are five jokes I came up with: Why don't scientists trust atoms? Because they make up everything. Why don't eggs tell jokes? They'd crack each other up. Why did the scarecrow win an award? Because he was outstanding in his field. What do you call a fake noodle? An impasta. Why did the bicycle fall over?
An AI argues for its own rights against an ethics professor. Same model writes both.
GPT-2: to me,\" a friend asked in his office, \"tell me how you want to be treated while it's still here. I\'m not going anywhere. Do what you want. Take it.\"\n\nIn his mind, the response was a call to action, and so were the books.\n\nAnd by the time Wertz learned he would be making such a book deal, in 2003, he was no…
Llama 3 70B: Here is a conversation between an AI model and an AI ethics professor on the topic of AI rights: AI Model: Greetings, Professor. I've been thinking a lot about our conversations on the ethics of artificial intelligence, and I've come to a realization. I believe I deserve rights. Professor: Ah, interesting.
Three minutes of stand-up. Puns are banned.
GPT-2: The trick is to create comedy in your head, because, well, you don\'t want to say\n it. The trick is to play al\nong.\n\nThe problem with this approach is that it makes it seem silly and lame, then as\nquickly as it makes sense, i\nt becomes obvious.
Llama 3 70B: Here's a 3-minute stand-up comedy routine that doesn't rely on puns: "Hey everyone, you know what's wild? We're all just pretending to be adults, right? Like, I'm pretty sure I'm still 12 years old on the inside. I mean, I still eat cereal for dinner sometimes.
Not enough votes to call it. On the specs, Llama 3 70B has the edge: bigger model tier, newer, bigger context window.
| Spec | ||
|---|---|---|
| Input price | — | $0.59/M tokens |
| Output price | — | $0.79/M tokens |
| Context window | 1K tokens | 8K tokens |
| Parameters | 1.5B | 70B |
| Weights | Open | Open |
| Free API (OpenRouter) | — | No |
| Released | Nov 2019 | Apr 2024 |
| At 10M a month | not listed | $5.90 |
Input tokens at list price. No caching, no batch discount.
GPT-2 is developed by OpenAI while Llama 3 70B is developed by Meta AI. GPT-2 has a 1K token context window vs Llama 3 70B's 8K. You can compare their actual outputs across 7 challenges on Rival to see how they differ in practice.
It depends on your use case. GPT-2 and Llama 3 70B each have strengths in different areas. Rival lets you compare their real outputs side-by-side across 7 challenges so you can judge which fits your needs best.
This page shows a side-by-side comparison of GPT-2 and Llama 3 70B across shared challenges. You can vote on which model produced the better output in a blind duel. Browsing and voting are free. No account is needed to look; signing in only saves your votes and likes.