Undoubtedly, the response generated by GPT-4 is impressive, but it has a limited access. I experimented with the paid version of GPT-3.5 and found that while it generated text quickly, the quality of the content was lacking. Surprisingly, the free version of GPT-3.5 produced better content than the paid version. Have others experienced this as well? I’ve attempted to use various prompts, but the paid version of GPT-3.5 consistently falls short in terms of quality.
I’ve experienced some differences between the snapshot models and the ‘latest’. It’s very hard to tell if that is random or not, however.
In order to have an informed opinion about this with proper sampling, I’d have to run model evals against the different endpoints. Sadly, I don’t have the time for that, but I think if someone took up this task they’d get a lot of eyeballs and traffic as this issue is something I see repeatedly posted everywhere.
That said, not entirely sure what the TOS is on posting evals publicly of OpenAI models. Does anyone know? Plus I wouldn’t want to upset the OpenAI team regardless of what the TOS says.
This seems to be the relevant section:
(c) Restrictions . You may not (i) use the Services in a way that infringes, misappropriates or violates any person’s rights; (ii) reverse assemble, reverse compile, decompile, translate or otherwise attempt to discover the source code or underlying components of models, algorithms, and systems of the Services (except to the extent such restrictions are contrary to applicable law); (iii) use output from the Services to develop models that compete with OpenAI; (iv) except as permitted through the API, use any automated or programmatic method to extract data or output from the Services, including scraping, web harvesting, or web data extraction; (v) represent that output from the Services was human-generated when it is not or otherwise violate our Usage Policies; (vii) buy, sell, or transfer API keys without our prior consent; or (viii), send us any personal information of children under 13 or the applicable age of digital consent. You will comply with any rate limits and other requirements in our documentation. You may use Services only in geographies currently supported by OpenAI.
1 Like
My bad, I misread the topic. 
1 Like
I have thoroughly examined the outputs generated from the same prompts using both the free and paid versions of GPT-3.5. Despite multiple checks, I have found that the free version produces superior responses compared to the paid version.I believe it is valuable to express this opinion for the sake of improvement, so that OpenAI can look into it. While it is important to prioritize speedy responses, the quality of the output must not be compromised.
1, The free version maybe updated by openai at any time, but the api may not,so they won’t be the same
2, Add this in your system message parameter in api should give you a better response:
“systemMessage”: “You are ChatGPT, a large language model trained by OpenAI. Answer as concisely as possible”,
1 Like
Use the ChatGPT Free website for your tasks. You have unlimited prompts with that website. www.chatgptfree.ai
That’s interesting, ayasir135! I’ve heard others mention similar experiences with GPT-3.5 versions. OpenAI might prioritise speed for free users so that quality could be affected.