I agree with you. As a digital marketer and tech enthusiast, I use a few tools for research, content writing, graphic designing, coding, video editing, and audio editing. I will list all my tools along with their use cases and best features in this community soon.
The backends of a few models are similar, but my use cases are different.
By the way, I am also working on automation, AI agents, and IVRs. Right now, I am learning, but I plan to launch a few services in India for businesses.
Yeah, I have seen this too frequently. When I respond with âare you sureâ, it goes to searching mode and then goes like âyouâre rightâ blah blah. Iâm getting such responses frequently nowadays. Canât rely on this for 100% accuracy.
I think it does that to avoid using too many tokens for every user and every prompt.
The only way to avert this is to use a good prompt. Unlike what most people think, LLMs prefer descriptive language with details. You can prompt it to refer to certain sources, exclude unverifiable sources, etc.
I have used Plus and it is good, responsive and accurate. I was testing both the versions and I find 2300 INR a bit too steep.
Did anyone use the new âGo Plan" ? Any feedback would help!
ChatGPT is absolutely horrible for research unless you use the GPT 5 Pro. It hallucinates so bad, does not read between the lines of the context. Gemini and Claude is way better. Even perplexity is so bad for research
Not sure what went wrong with perplexity for you but it clearly shows sources where its getting info from. I find it to be the best one when doing research on anything.
I believe they are nerfing the models more and more these days, especially for research where it seems to spend less time now. There are workarounds but itâs a bit more noticeable lately.