What is your favourite LLM model for article writing?
- Thread starter CRISPRSeo
- Start date
- Status
ClintonORM
Newbie
- Nov 6, 2025
- 23
- 7
well prompted chatGPT is enough at my opinion
CRISPRSeo
Newbie
- Nov 10, 2025
- 24
- 8
Agree. I also use ChatGPT for other daily tasks.
Yavnekuralio
Junior Member
- Oct 2, 2025
- 105
- 22
bro ı stick with GPT-5 mini fast, reliable, and handles reasoning well. I like how flexible it is with different styles and prompts.
I'm surprised by how many ChatGPT responses there are here! I've tried many of the models, and almost every single time I've found the output to be extremely verbose and doesn't pass the "not AI" sniff test. Then you add in the whole laughable debacle "don't use em dashes" in the prompt. For me, even with really detailed prompts around styling, it never worked well.
On the other hand, Claude Sonnet always read more natural and required far less prompting.
On the other hand, Claude Sonnet always read more natural and required far less prompting.
ngomatshibanda
Newbie
- Oct 13, 2025
- 3
- 0
Well crafted grok prompts work perfectly fine
- Aug 4, 2015
- 7,021
- 883
My favorite model changes depending on the task, but I usually use GPT models because they’re fast, flexible, and good at long content
- Aug 16, 2014
- 3,086
- 680
Claude Sonnet 4.5 is solid. I usually prefer GPT models for creativity and heavy reasoning since they are fast, flexible, and handle complex prompts well. But the best choice always depends on what you are working on.
s3idi15
BANNED
- May 11, 2018
- 68
- 18
for me it kinda depends on what i'm working on. if i need long, detailed stuff or something a bit more “thoughtful,” i usually stick to claude too because it handles nuance really well. but for quick ideas, creative angles, or just getting something done fast without overthinking, i lean on chatgpt. feels smoother for day-to-day stuff. i’ve also started mixing models sometimes… like generating with one and refining with another. weirdly works pretty well.
CRISPRSeo
Newbie
- Nov 10, 2025
- 24
- 8
Interesting! I didn't try Mistral yet. I have to test it.
AndrewYaHnO
Newbie
- Aug 7, 2023
- 29
- 10
I usually stick with GPT-5.1 for most tasks
CRISPRSeo
Newbie
- Nov 10, 2025
- 24
- 8
I love Gemini 3.0 Pro, and I am using Nano Banana for creating visual graphics for my blog posts, as well as thumbnails for YouTube videos.
Agree dude. Em dashes are irritating in ChatGPT, although Sam Altman suggests using custom instructions to eliminate them. I see people in X, but it's not working. Claude is more natural, a big fan of Claude.
Fiducia
BANNED
- Mar 1, 2020
- 1,469
- 1,160
Give it a try! Works just great for German and French, way more natural than other LLMs trimmed on English language.
muhammadshumbul
Junior Member
- Oct 2, 2025
- 122
- 33
I’ve been using gbt5 mini too its surprisingly consistent and adapts well to complex prompts without overcomplicating things.
That's interesting, 4.5 is definitely strong on reasoning.
What kind of reasoning function are you applying? Are you mostly using it for content outlines or for generating full articles?
- Status