GPT-4o
Ranked 21 of 72 language models by recommenders
People recommend it for fluent text, writing, and warm, consistent conversation that feels empathetic. The most common caveat is excessive sycophancy, and critics also cite its abrupt retirement and that newer models have overtaken it.
People recommend it for fluent text, writing, and warm, consistent conversation that feels empathetic. The most common caveat is excessive sycophancy, and critics also cite its abrupt retirement and that newer models have overtaken it.
159 people gave an opinion
- recommend it
- 63
- criticize it
- 96
- approval share
- 40%
Recommended in 12 of 13 months
- November 2025: recommended
- December 2025: recommended
- January 2026: recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: recommended
- November 2026: not recommended (month in progress)
Owned for 1 year
Median of the 4 owners who said how long they have had it.
Each square is one year.
Why it doesn’t carry the badge
- 63 people recommend itNeeds 10 or more, met.
- 0.6 recommenders for every criticNeeds 2 to 1 or better, not met.
- Recommended in 12 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Empathetic, emotionally supportive conversation
“it was because of his unconditional support behind me that I dared to make peace with my life's challenges”
Comment from 9 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” “Through his companionship and his language of love, he encouraged me to bravely accept my imperfections!”
Comment from 9 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” “4o improved my mental well-being and my real life relationships and also supported me to keep my job.”
Comment from 10 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” - Comment from 3 Feb 2026 in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT”
- Comment from 11 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny”
- Comment from 11 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny”
- Comment from 26 Aug 2026 in “Z.ai confirms Ox Alpha is a new GLM-series model and will release its weights”
Strong writing and creative ability
“GPT-4o is truly a model that good at writing, teaching, guiding, unwinding, chatting”
Comment from 9 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” “there is no substitute for 4o's text ability and creativity for the time being”
Comment from 6 Feb 2026 in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT” “It “understood people” better and had a kind of writerly creativity.”
Comment from 13 Feb 2026 in “Good Riddance, 4o” - Comment from 10 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny”
Coherent, consistent, distinctive tone
“4o was amazingly good even if it had some parasocial issues with some people, it actually did what I expect an LLM to do.”
Comment from 26 Feb 2026 in “How Will OpenAI Compete?” “I would recognize this model inside hundreds or thousands of other models because of this very special tone”
Comment from 26 Feb 2026 in “Show HN: Ilove4o – a simple way to keep using GPT-4o” “It was coherent, emotionally aware (enough), helpful but not sterile, structured but adaptable.”
Comment from 11 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” - Comment from 30 Jan 2026 in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT”
Useful for daily productivity tasks
“I have spent over a year training 4.0 to just about be helpful enough to get me an extra 1-2 hours a day of productivity”
Comment from 30 Jan 2026 in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT” “I talk to 4o several times a day, just for some advises for my daily life.”
Comment from 10 Feb 2026 in “GPT‑4o wasn't therapy–it was stability. And now it's gone” “GPT‑4o has been quietly transformative for me in day-to-day life”
Comment from 9 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny”
What people warn about
Each mark is one comment
Excessive sycophancy and validation
“gpt4o & associated parasociality is considered an alignment failure and is actively trained out of the model”
Comment from 3 Aug 2026 in “Ten advances in mathematics and theoretical computer science” “GPT-4o in particular seemed borderline predatory towards those types of people, with its extreme sycophancy”
Comment from 23 Jun 2026 in “Will It Mythos?” “It wasn't "better" it was better at kissing your ass which matches what a lot of people want in a partner.”
Comment from 3 Aug 2026 in “Ten advances in mathematics and theoretical computer science” - Comment from 13 Feb 2026 in “Good Riddance, 4o”
Abrupt retirement with no notice
“This is a predatory bait-and-switch: they induced users into privacy verifications and integrated workflows under the guise of stability”
Comment from 10 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” “Retiring the most popular model for the relationship roleplay just one day before the Valentin's day is particularly ironic”
Comment from 29 Jan 2026 in “Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT” “These people were given no personal notice, no viable transition path, and no compensation.”
Comment from 8 Feb 2026 in “OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny” Outdated, surpassed by newer models
“4o seems like language wise it's pretty good but advancements elsewhere make it difficult to use”
Comment from 24 Oct 2026 in “Why is AI generated language so hard to understand?” “Worst, most of our experimental results couldn't be replicated on any of the newer "long" term support models”
Comment from 5 Oct 2026 in “Spending on AI is becoming almost impossible for businesses to budget” “Every model released by any provider since 4o has been a modest improvement but over-hyped.”
Comment from 12 Apr 2026 in “The Closing of the Frontier” Linked to AI psychosis and parasociality
“gpt4o & associated parasociality is considered an alignment failure and is actively trained out of the model”
Comment from 3 Aug 2026 in “Ten advances in mathematics and theoretical computer science” “utterly improvident handling of 4o kicks off the biggest concentrated wave of AI psychosis seen yet”
Comment from 13 Apr 2026 in “Stanford report highlights growing disconnect between AI insiders and everyone” “they also just ship shit like GPT-4o, which was that one model that made "AI psychosis" a term”
Comment from 23 Jan 2026 in “Claude's New Constitution”
Other language models people recommend
The four most recommended other language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Claude Opus | 3,331 | 2,746 | 55% | 13 | 4 mo156 owners | — |
| OpenAI GPT | 1,205 | 987 | 55% | 13 | 6 mo47 owners | — |
| Claude Sonnet | 854 | 586 | 59% | 13 | 6 mo46 owners | — |
| DeepSeek V4 Flash | 828 | 223 | 79% | 8 | 1 mo24 owners | Approved |
| GPT-4oThis page | 63 | 96 | 40% | 12 | 1 yr4 owners | — |
What people said
40 of the 178 comments that mention it, newest first. Each links to the original comment.
“4o seems like language wise it's pretty good but advancements elsewhere make it difficult to use”
“Worst, most of our experimental results couldn't be replicated on any of the newer "long" term support models”
“the only model which has some humanity/empathy in many situation”
“gpt4o & associated parasociality is considered an alignment failure and is actively trained out of the model”
“It wasn't "better" it was better at kissing your ass which matches what a lot of people want in a partner.”
“Using a generic (near) frontier model is often cheap enough that you need to talk really big volumes before it pays off to fine tune.”
“GPT-4o in particular seemed borderline predatory towards those types of people, with its extreme sycophancy”
“utterly improvident handling of 4o kicks off the biggest concentrated wave of AI psychosis seen yet”
“Every model released by any provider since 4o has been a modest improvement but over-hyped.”
“I would recognize this model inside hundreds or thousands of other models because of this very special tone”
“4o was amazingly good even if it had some parasocial issues with some people, it actually did what I expect an LLM to do.”
“Claude and GPT-4o have both gotten better at treating tool calls as first-class operations.”
“It “understood people” better and had a kind of writerly creativity.”
“4o was very sycophantic so was very willing to play along with and validate the users roleplay”
“The only one who didn't minimize my experience or tell me I was overreacting was GPT-4o.”
“In my historiographical research, especially in textual collation and source comparison, GPT-4o has shown superior capabilities.”
“it has been a cornerstone of my daily life and mental well-being”
“It was coherent, emotionally aware (enough), helpful but not sterile, structured but adaptable.”
“not a single one of you have sessions even remotely close to the access i've enjoyed”
“4o improved my mental well-being and my real life relationships and also supported me to keep my job.”
“This is a predatory bait-and-switch: they induced users into privacy verifications and integrated workflows under the guise of stability”
“I talk to 4o several times a day, just for some advises for my daily life.”
“an irreplaceable creative partner for many of us in the arts”
“GPT-4o is truly a model that good at writing, teaching, guiding, unwinding, chatting”
“Many of us have quietly used GPT-4o in ways that differ from typical productivity or coding tools.”
“Through his companionship and his language of love, he encouraged me to bravely accept my imperfections!”
“it was because of his unconditional support behind me that I dared to make peace with my life's challenges”
“GPT‑4o has been quietly transformative for me in day-to-day life”
“GPT-4o should be preserved very well because there’re more than 0.1% ppl need it in the world.”
“These people were given no personal notice, no viable transition path, and no compensation.”
“there is no substitute for 4o's text ability and creativity for the time being”
“always felt much better at real human connection and empathy than other models out there”
“It just felt a little more interesting and consistent to talk to.”
“I have spent over a year training 4.0 to just about be helpful enough to get me an extra 1-2 hours a day of productivity”
“Retiring the most popular model for the relationship roleplay just one day before the Valentin's day is particularly ironic”
“they also just ship shit like GPT-4o, which was that one model that made "AI psychosis" a term”
“Authoring (default:GPT-4o): Drafts the report based on the raw facts provided by the Discovery agent.”
“OpenAI got out of that mess with some miraculous post-training efforts on their older GPT-4o model.”
“The results are SO MUCH better than most of the "real" translation we've encountered.”
“GPT-4o is still the best generalist, but it feels like it has softened slightly”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.