Hello! Wanted to ask, what Plugin would you recommend for literature research on ChatGPT? I tried out a few, but im not really conviced yet. Im working in the field of biology, specifically agriculture!
Im aware of the general function of ChatGPT (that it does not provide real sources), but i thought since there is an option for the Plugins (and they do provide sources, just not the most accurate ones…), i may ask about your experience 
Hi and welcome to the community!
I’ve updated your title to better reflect what you’re looking for and help people with the relevant knowledge find it. I also added a couple of relevant tags for the same reason.
There are many ChatGPT plugins for accessing online research papers. In my personal experience, they have all been pretty bad at properly parsing and understanding math display environments in PDFs which is just about the only thing that matters to me as Statistics PhD student, so I haven’t really used them beyond a cursory inspection. I’m currently rolling my own yet-another-other-chat-with-your-pdf-file plugin for personal use.
Perhaps you could expand on what specific issues you’ve encountered related to your field of research that leave you unimpressed and someone else will be able to better guide you to a solution.
Here is an example, YMMV
ChatGPT July 20 Version
Model: Plugins
Enabled plugins: Science
Prompt
Give me a list of recent agriculture papers.
Completion
Here are some recent papers related to agriculture:
L’Agriculture biologique, une innovation territoriale au service du développement rural : le cas du Gers by Charlène Arnaud and Pierre Triboulet, published in 2025. The paper discusses organic farming as a territorial innovation that serves rural development, with a case study on Gers, the leading “organic department” in France.
Factors affecting Vietnamese agricultural commodities’ export to the EU market by Tran Nho Quyet, Nguyen Minh Ngoc, Nguyen Hoang Tien, and Le Doan Minh Duc, published in 2025. The abstract for this paper is not available.
Isotopic characteristics and source analysis of atmospheric ammonia during agricultural periods in the Xichuan area of the Danjiangkou Reservoir by Xiaoshu Chen, Ting Zhao, Chunyan Xiao, Xiaoming Gao, and Feihong Chen, published in 2024. The paper discusses nitrogen deposition as an important means of exogenous nitrogen input in reservoir water, with a focus on the impact of agricultural activities.
Would you like to know more about any of these papers?
It’s quite intriguing to notice that all the publish dates are placed in the future. This observation leads us to wonder if there might be a lack of agriculture papers in the mix.
Thank you for the recommendation! What do you think is the advantage of that tool compared to e.g. ScholarAI (which didn’t really impress me in the first tries …)
elmstedt
Thanks for your reply and for the valuable adjustments! I try to be a bit more precise in my next posts!
Well, in general, the papers did not really connect with the topic. e.g. I wanted to know about agro-ecological conditions in a specific country, which shouldn’t be a too difficult question, imo.
The suggested paper were kind of random, advising me about specific plants and so on.
I also tried to ask first the question in general and then asked to provide me sources for the provided information, worked a bit better, but not significantly.
Tried out many plugins, but was somehow a bit overwhelmed by the variety of tools there … So i asked, if someone would have specific recommendations.
I assume, no Plugin is working with Elsevier because of the access restrictions?
I don’t wanna complain here, i’ve been really happy with everything. Just tryna search for solutions!
Don’t use ScholarAI because Science does what I need. However I don’t use Science that often. Find that Arxiv queries like this work better for what I seek as I scan the list daily.
https://arxiv.org/list/cs.SC/recent
Today on a similar list found something related to this reply.
Thank you, Eric! I will definitely check that tool out! Do you have any recommendations for the search string? Or is that not highly important?
Thanks a lot for your recommendations and advices!
I see! This is strange, i agree. There definitely may be a lack of agricultural papers. Maybe considering the journal, i normally do my research in Elsevier and CAB
You might find that using this site to find the agriculture papers first is better
https://www.cabidigitallibrary.org/journal/agrirxiv
then using one of the plugins to chat with the paper of value.
To find the chat with PDF type plugins
pdf in the search box.Today there are 21 such plugins.
Sorry, you mean, searching the paper on the website itself myself and then summarizing the value with a Plugin? yes, i tried that! That works pretty decent
Can you also clarify for me, are you looking to identify relevant papers, interact with them in a chat interface, or both?
Hey, im looking to identify relevant papers as a first step! I found already a good Plug-in to work with them in a chat interface, at least I think it’s pretty solid. But the identification part is still a lil hard 
I’m interested in tinkering with this, can you give me 1–3 specific prompts you might use for your research?
Sure! So three examples would be:
Can you provide me three papers about the agroecological conditions in Argentina?
Please provide me a paper that explain the history of cattle-farming in Brazil.
Can you provide me a paper that evaluates the impact of cattle-farming on the environment in Brazil?
That would be some interesting examples, the specific locations (Argentina, Brazil …) can be exchangeable
I’m working on an experimental process now for surfacing relevant literature, I tried it out on your third example and this is what it came up with.
DOI:10.1007/s11367-016-1062-4
Environmental Impact assessment of a semi-intensive beef cattle production in Brazil’s Northeast
C. Willers, Henrique Leonardo Maranduba, +1 author L. Rodrigues
Published 1 April 2017
The International Journal of Life Cycle Assessment
DOI:10.34117/bjdv6n11-105
ANÁLISE DOS FATORES QUE CAUSAM IMPACTO AMBIENTAL NA PRODUÇÃO DE BOVINOCULTURA DE CORTE / ANALYSIS OF FACTORS THAT CAUSE ENVIRONMENTAL IMPACT ON THE PRODUCTION OF CUTTING CATTLE
Ramon Luan Pereira da Silva, Lilian Natália Ferreira Lima, +5 authors Renilda Silva Soares
Published 2020
Brazilian Journal of Development
DOI:10.1007/s10457-019-00460-x
Silvopastoral management of beef cattle production for neutralizing the environmental impact of enteric methane emission
Leonardo Resende, Marcelo Dias Müller, +4 authors L. Rego
Published 7 November 2019
Agroforestry Systems
DOI:10.1038/s41598-019-47647-x
Biochar amendment improves degraded pasturelands in Brazil: environmental and cost-benefit analysis
A. Latawiec, B. Strassburg, +16 authors S. Hale
Published 19 August 2019
Scientific Reports
DOI:10.1093/jas/skab235.771
PSVIII-19 Greenhouse gas emissions from beef cattle production in Brazil: how we can mitigate from animal operations?
Andre Pastori D Aurea, A. Cardoso, +3 authors Yuri Santa Rosa Guimarães
Published 8 October 2021
Journal of Animal Science
As this is well outside my area of expertise I don’t know if thiese are quite the type of result you’d be hoping to surface. To my lay understanding they don’t seem terrible.
I think I need to do some tweaking to weight results by journal and author reputation, maybe citation count, etc in order to preferentially surface more significant papers then also somehow balance that with recency.
Maybe find the most important papers relevant to a query, then find papers which cite many of the most important papers, filter to the last 3–5 years, then select some subset of those most relevant to the original query, then sort that list by summer measure of importance based on journal, author, and citation count…
I think there’s a lot to be done here because I think most of the existing plugins are doing fairly naive searches.
Sharing the case of a friend who was looking for the original paper to link ovarian cancer to ADAM17.
Prompt: search the science plug-in for the oldest paper regarding ADAM17 in conjunction with ovarian cancer
The model reply was irrelevant unfortunately.
In addition to your concept to identifying the most relevant papers from a set of possible correct answers, I am a strong proponent of querying the user to help the model fulfill the request. Especially in a multi-step look-up, compare, analyze scenario like this one.
In many other contexts I’d say that a single better prompt could suffice but in this scenario I don’t see it happening unless the model can access a specific dataset for Gynecology/Cancer with the user supporting the model to evaluate the results.
Looking for “the original paper” is a typical use case but it’s also time consuming, if not challenging and often our specialists at the library handle these searches. If I recall correctly you are working with a University? They may be able to provide more specific pointers for effective search strategies.
PS. My function to reply to specific comments has stopped working. I am referring to @elmstedt . Sorry for the ping.
Hey, I agree here with you, the results are pretty good! Is already quite the step from what I saw in the results of my first plugin try-outs! It is not 100 % precise, but definitely solid! Respect, cool!
I completely agree with your strategy to improve the accuracy of the model-searches! Sounded pretty solid thought-off
Indeed, I don’t think the search strategies of currently models were comprehensively and well thought-off, rather naive
Greetings
Sounds great! I don’t have the technological knowledge in this field of coding a model, I think it is great when on the first search some pretty solid results are presented. Then you could interact and specify which papers of the presented ones are maybe irrelevant (because of reason X)
It was not so much about looking for a original paper of a source that I found somewhere, rather it was about looking for a source that underlines the information ChatGPT was giving me at a first prompt
Kind of like that:
What impact will climate change have on extensive cattle husbandry systems in Latin-America?
…
Can you provide me papers for that topic?
…
Kind of like that … am I understandable? Maybe I also expect to much from the model, I’m not sure, but the results @anon22939549 showed, were definitely a great improvement and maybe shows potential?