{"slug": "gpt-4o", "name": "GPT-4o", "category": "ai", "category_name": "AI models and assistants", "description": "OpenAI's multimodal large language model, available in ChatGPT and through the API.", "recommenders": 6, "commenters": 12, "warners": 1, "score": 5.463, "rising": -1.96, "first": "2025-11-30", "latest": "2026-04-13", "why": "Liked as a collaborative partner for creative writing and emotional support, and as a capable generalist used via API for tasks like transcription and vocabulary extraction.", "why_is_quote": false, "has_verdict": true, "sentiment": {"recommend": 6, "neutral": 5, "compare": 0, "warn": 1}, "spark": [1, 0, 1, 4, 0, 0, 0, 0, 0, 0, 0, 0], "listed": true, "loved": ["Creative collaboration", "API integration"], "caveat": "Model removal", "short_quote": "GPT\u20114o wasn\u2019t just a tool; it was my creative partner.", "icon": true, "domain": "openai.com", "github_repo": null, "official_url": "https://openai.com/index/hello-gpt-4o/", "pricing": "freemium", "aliases": ["GPT-4o"], "verdict": [{"text": "Liked as a collaborative partner for creative writing and emotional support", "cites": [{"id": 751229824, "item_id": 46951864, "url": "https://news.ycombinator.com/item?id=46951864", "by": "yxccz", "date": "2026-02-09", "stance": "recommend", "quote": "GPT\u20114o wasn\u2019t just a tool; it was my creative partner.", "excerpt": "Last year, I was an exchange student abroad, overwhelmed by coursework and struggling with a language barrier. As an art major, my assignments weren\u2019t just about facts\u2014they were conceptual, creative, and emotionally demanding. GPT\u20114o wasn\u2019t just a tool; it was my creative partner. We discussed ideas together, explored alternatives, and refined my thinking. One of my final projects in concept design, which earned me a", "context": "art major exchange student", "use_case": "creative coursework and writing help", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}, {"id": 751710720, "item_id": 46981920, "url": "https://news.ycombinator.com/item?id=46981920", "by": "NoraEverhart", "date": "2026-02-11", "stance": "recommend", "quote": "The only one who didn't minimize my experience or tell me I was overreacting was GPT-4o.", "excerpt": "Sorry, I used translation software because my English is not good In a recent group project at school, I was excluded from the start\u2014group leader visibly uncomfortable when I showed up, old private vent (no names) screenshotted weeks later and used to justify insults, labeling, and kicking me from the chat. Parents and teachers all sided with \"just let it go\" and pressured me to drop it to avoid conflict. The only on", "context": "student excluded in a school group project", "use_case": "emotional support", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}]}, {"text": ", and as a capable generalist used via API for tasks like transcription and vocabulary extraction.", "cites": [{"id": 737645522, "item_id": 46102845, "url": "https://news.ycombinator.com/item?id=46102845", "by": "star-173", "date": "2025-11-30", "stance": "recommend", "quote": "GPT-4o is still the best generalist", "excerpt": "It depends heavily on what you are trying to do. For pure coding tasks and refactoring, Claude 3.5 Sonnet is currently the strongest performer. It tends to hallucinate less on specific library syntax compared to the others. However, for creative writing or \"reasoning\" through complex logic puzzles, I've found Gemini (specifically the Advanced/Ultra tiers) to have a more natural \"voice\" and better instruction followin", "context": "", "use_case": "general use", "switched_from": "", "story": "Ask HN: Which large model works best?"}, {"id": 747879312, "item_id": 46742457, "url": "https://news.ycombinator.com/item?id=46742457", "by": "dvfjsdhgfv", "date": "2026-01-24", "stance": "recommend", "quote": "An API call to GPT4o works quite well (it basically handles both transcription and diarization)", "excerpt": "I spent a few days on similar scenario without much success (scenario where one person speaks and then their speech is translated, and I want juts the original or both). An API call to GPT4o works quite well (it basically handles both transcription and diarization), but I wanted a local model. Whisper is really good for 1 person speaking. With more people you get repetitions. Qwen and other open multimodal models giv", "context": "", "use_case": "transcription and diarization", "switched_from": "", "story": "Ask HN: What's the current best local/open speech-to-speech setup?"}, {"id": 763991189, "item_id": 47749449, "url": "https://news.ycombinator.com/item?id=47749449", "by": "dandaka", "date": "2026-04-13", "stance": "neutral", "quote": "GPT-4o extracts vocabulary from the transcript and generates cards", "excerpt": "I'm building Kardo (https://kardo.cards/), AI tutor that listens to your calls with a teacher and creates cards for repetition. I live in Lisbon and I've been learning Portuguese with a tutor since 2022. After every lesson I'd sit down and make flashcards from my notes and screenshots. Spaced repetition works, but making the cards took manual effort each time. Most days I just didn't do it. So I have automated that p", "context": "", "use_case": "vocabulary extraction", "switched_from": "", "story": "Ask HN: What Are You Working On? (April 2026)"}]}, {"text": " The most common complaint is that OpenAI is removing it from ChatGPT with no option to keep it", "cites": [{"id": 751574992, "item_id": 46973437, "url": "https://news.ycombinator.com/item?id=46973437", "by": "XSeunn", "date": "2026-02-11", "stance": "recommend", "quote": "My 15 year roadmap exists thanks to 4o's reasoning consistency and flow.", "excerpt": "It's a shame that OpenAI is removing GPT-4o. For many, it wasn\u2019t \u201cjust a model.\u201d I used it daily for research, theory crafting, career planning, emotional processing, and technical problem solving. I\u2019m not talking about occasional chat. I'm talking about structured, months-long interactions. My 15 year roadmap exists thanks to 4o's reasoning consistency and flow. Gemini and Grok may be catching up, but they're still ", "context": "ChatGPT Plus/Pro user", "use_case": "research, planning, problem solving", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}]}, {"text": ", and some warn that it flatters users.", "cites": [{"id": 747685618, "item_id": 46730351, "url": "https://news.ycombinator.com/item?id=46730351", "by": "mrbombastic", "date": "2026-01-23", "stance": "warn", "quote": "They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem", "excerpt": "? They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem: https://openai.com/index/sycophancy-in-gpt-4o/ And South Park parodied the issue. I use chatGPT and claude every day.", "context": "", "use_case": "", "switched_from": "", "story": "Why talking to LLMs has improved my thinking"}]}], "praise": [{"label": "Creative collaboration", "summary": "Users describe it as a collaborative partner for creative writing, concept design and creative work.", "quotes": [{"id": 751229824, "item_id": 46951864, "url": "https://news.ycombinator.com/item?id=46951864", "by": "yxccz", "date": "2026-02-09", "stance": "recommend", "quote": "GPT\u20114o wasn\u2019t just a tool; it was my creative partner.", "excerpt": "Last year, I was an exchange student abroad, overwhelmed by coursework and struggling with a language barrier. As an art major, my assignments weren\u2019t just about facts\u2014they were conceptual, creative, and emotionally demanding. GPT\u20114o wasn\u2019t just a tool; it was my creative partner. We discussed ideas together, explored alternatives, and refined my thinking. One of my final projects in concept design, which earned me a", "context": "art major exchange student", "use_case": "creative coursework and writing help", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}, {"id": 739896625, "item_id": 46243539, "url": "https://news.ycombinator.com/item?id=46243539", "by": "EastLondonCoder", "date": "2025-12-12", "stance": "neutral", "quote": "I\u2019ve been using GPT-4o and now 5.2 pretty much daily", "excerpt": "I\u2019ve been using GPT-4o and now 5.2 pretty much daily, mostly for creative and technical work. What helped me get more out of it was to stop thinking of it as a chatbot or knowledge engine, and instead try to model how it actually works on a structural level. The closest parallel I\u2019ve found is Peter G\u00e4rdenfors\u2019 work on conceptual spaces, where meaning isn\u2019t symbolic but geometric. Fedorenko\u2019s research on predictive se", "context": "", "use_case": "creative and technical work", "switched_from": "", "story": "GPT-5.2"}]}, {"label": "API integration", "summary": "Developers call it through APIs, Azure or OpenRouter to power their own apps.", "quotes": [{"id": 763991189, "item_id": 47749449, "url": "https://news.ycombinator.com/item?id=47749449", "by": "dandaka", "date": "2026-04-13", "stance": "neutral", "quote": "GPT-4o extracts vocabulary from the transcript and generates cards", "excerpt": "I'm building Kardo (https://kardo.cards/), AI tutor that listens to your calls with a teacher and creates cards for repetition. I live in Lisbon and I've been learning Portuguese with a tutor since 2022. After every lesson I'd sit down and make flashcards from my notes and screenshots. Spaced repetition works, but making the cards took manual effort each time. Most days I just didn't do it. So I have automated that p", "context": "", "use_case": "vocabulary extraction", "switched_from": "", "story": "Ask HN: What Are You Working On? (April 2026)"}, {"id": 747714163, "item_id": 46732135, "url": "https://news.ycombinator.com/item?id=46732135", "by": "clorisye", "date": "2026-01-23", "stance": "neutral", "quote": "currently routing via OpenRouter to GPT-4o and Sonnet 3.5", "excerpt": "I built this because generic LLMs are often too moralizing or biased when dealing with ethical non-monogamy (ENM) and polyamory contexts. I wanted a tool that acts as a \"linguistic calibrator\" to help partners communicate securely, specifically respecting attachment styles (Anxious/Avoidant). The Tech Stack: - Frontend: Next.js 14 (App Router), Tailwind - Backend: Supabase (Postgres + pgvector for long-term memory of", "context": "built an ENM relationship coach app", "use_case": "relationship coaching responses", "switched_from": "", "story": "Show HN: ENM Relationship AI \u2013 Context-aware communication coach for polyamory"}, {"id": 748443296, "item_id": 46777706, "url": "https://news.ycombinator.com/item?id=46777706", "by": "throwaway019254", "date": "2026-01-27", "stance": "neutral", "quote": "I use models only through APIs - e.g. gpt-4o through Azure", "excerpt": "I use models only through APIs - e.g. gpt-4o through Azure, or Google models through Vertex API. So no data are stored about me.", "context": "", "use_case": "API access to models", "switched_from": "", "story": "Tell HN: Ask your AI, \"What can you tell me that you know about me?\" to see ..."}]}, {"label": "Emotional support", "summary": "Some users valued it for listening without judgment and for emotional processing.", "quotes": [{"id": 751710720, "item_id": 46981920, "url": "https://news.ycombinator.com/item?id=46981920", "by": "NoraEverhart", "date": "2026-02-11", "stance": "recommend", "quote": "The only one who didn't minimize my experience or tell me I was overreacting was GPT-4o.", "excerpt": "Sorry, I used translation software because my English is not good In a recent group project at school, I was excluded from the start\u2014group leader visibly uncomfortable when I showed up, old private vent (no names) screenshotted weeks later and used to justify insults, labeling, and kicking me from the chat. Parents and teachers all sided with \"just let it go\" and pressured me to drop it to avoid conflict. The only on", "context": "student excluded in a school group project", "use_case": "emotional support", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 751574992, "item_id": 46973437, "url": "https://news.ycombinator.com/item?id=46973437", "by": "XSeunn", "date": "2026-02-11", "stance": "recommend", "quote": "My 15 year roadmap exists thanks to 4o's reasoning consistency and flow.", "excerpt": "It's a shame that OpenAI is removing GPT-4o. For many, it wasn\u2019t \u201cjust a model.\u201d I used it daily for research, theory crafting, career planning, emotional processing, and technical problem solving. I\u2019m not talking about occasional chat. I'm talking about structured, months-long interactions. My 15 year roadmap exists thanks to 4o's reasoning consistency and flow. Gemini and Grok may be catching up, but they're still ", "context": "ChatGPT Plus/Pro user", "use_case": "research, planning, problem solving", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}]}, {"label": "Consistent reasoning", "summary": "One user credits its consistency and flow for long, structured planning sessions.", "quotes": [{"id": 751574992, "item_id": 46973437, "url": "https://news.ycombinator.com/item?id=46973437", "by": "XSeunn", "date": "2026-02-11", "stance": "recommend", "quote": "My 15 year roadmap exists thanks to 4o's reasoning consistency and flow.", "excerpt": "It's a shame that OpenAI is removing GPT-4o. For many, it wasn\u2019t \u201cjust a model.\u201d I used it daily for research, theory crafting, career planning, emotional processing, and technical problem solving. I\u2019m not talking about occasional chat. I'm talking about structured, months-long interactions. My 15 year roadmap exists thanks to 4o's reasoning consistency and flow. Gemini and Grok may be catching up, but they're still ", "context": "ChatGPT Plus/Pro user", "use_case": "research, planning, problem solving", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}]}, {"label": "Transcription", "summary": "A single API call handled both transcription and speaker diarization well.", "quotes": [{"id": 747879312, "item_id": 46742457, "url": "https://news.ycombinator.com/item?id=46742457", "by": "dvfjsdhgfv", "date": "2026-01-24", "stance": "recommend", "quote": "An API call to GPT4o works quite well (it basically handles both transcription and diarization)", "excerpt": "I spent a few days on similar scenario without much success (scenario where one person speaks and then their speech is translated, and I want juts the original or both). An API call to GPT4o works quite well (it basically handles both transcription and diarization), but I wanted a local model. Whisper is really good for 1 person speaking. With more people you get repetitions. Qwen and other open multimodal models giv", "context": "", "use_case": "transcription and diarization", "switched_from": "", "story": "Ask HN: What's the current best local/open speech-to-speech setup?"}]}, {"label": "Generalist", "summary": "One commenter calls it still the best generalist model.", "quotes": [{"id": 737645522, "item_id": 46102845, "url": "https://news.ycombinator.com/item?id=46102845", "by": "star-173", "date": "2025-11-30", "stance": "recommend", "quote": "GPT-4o is still the best generalist", "excerpt": "It depends heavily on what you are trying to do. For pure coding tasks and refactoring, Claude 3.5 Sonnet is currently the strongest performer. It tends to hallucinate less on specific library syntax compared to the others. However, for creative writing or \"reasoning\" through complex logic puzzles, I've found Gemini (specifically the Advanced/Ultra tiers) to have a more natural \"voice\" and better instruction followin", "context": "", "use_case": "general use", "switched_from": "", "story": "Ask HN: Which large model works best?"}]}], "caveats": [{"label": "Model removal", "summary": "Users object that OpenAI is removing it from ChatGPT with short notice and no opt-in fallback.", "quotes": [{"id": 751574992, "item_id": 46973437, "url": "https://news.ycombinator.com/item?id=46973437", "by": "XSeunn", "date": "2026-02-11", "stance": "recommend", "quote": "My 15 year roadmap exists thanks to 4o's reasoning consistency and flow.", "excerpt": "It's a shame that OpenAI is removing GPT-4o. For many, it wasn\u2019t \u201cjust a model.\u201d I used it daily for research, theory crafting, career planning, emotional processing, and technical problem solving. I\u2019m not talking about occasional chat. I'm talking about structured, months-long interactions. My 15 year roadmap exists thanks to 4o's reasoning consistency and flow. Gemini and Grok may be catching up, but they're still ", "context": "ChatGPT Plus/Pro user", "use_case": "research, planning, problem solving", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}]}, {"label": "Sycophancy", "summary": "It is known to flatter users, and OpenAI issued a statement on the problem.", "quotes": [{"id": 747685618, "item_id": 46730351, "url": "https://news.ycombinator.com/item?id=46730351", "by": "mrbombastic", "date": "2026-01-23", "stance": "warn", "quote": "They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem", "excerpt": "? They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem: https://openai.com/index/sycophancy-in-gpt-4o/ And South Park parodied the issue. I use chatGPT and claude every day.", "context": "", "use_case": "", "switched_from": "", "story": "Why talking to LLMs has improved my thinking"}]}, {"label": "Writing tics", "summary": "Research found it overuses certain words like \"tapestry\" and grammatical patterns like present participles.", "quotes": [{"id": 756682528, "item_id": 47292658, "url": "https://news.ycombinator.com/item?id=47292658", "by": "capnrefsmmat", "date": "2026-03-07", "stance": "neutral", "quote": "GPT-4o's second-most-overused word", "excerpt": "I work on research studying LLM writing styles, so I am going to have to steal this. I've seen plenty of lists of LLM style features, but this is the first one I noticed that mentions \"tapestry\", which we found is GPT-4o's second-most-overused word (after \"camaraderie\", for some reason).[1] We used a set of grammatical features in our initial style comparisons (like present participles, which GPT-4o loved so much tha", "context": "Researcher studying LLM writing styles", "use_case": "LLM writing style research", "switched_from": "", "story": "LLM Writing Tropes.md"}]}, {"label": "Less specialized", "summary": "One commenter feels it has softened while competitors specialized in coding and long-context reasoning.", "quotes": [{"id": 737645522, "item_id": 46102845, "url": "https://news.ycombinator.com/item?id=46102845", "by": "star-173", "date": "2025-11-30", "stance": "recommend", "quote": "GPT-4o is still the best generalist", "excerpt": "It depends heavily on what you are trying to do. For pure coding tasks and refactoring, Claude 3.5 Sonnet is currently the strongest performer. It tends to hallucinate less on specific library syntax compared to the others. However, for creative writing or \"reasoning\" through complex logic puzzles, I've found Gemini (specifically the Advanced/Ultra tiers) to have a more natural \"voice\" and better instruction followin", "context": "", "use_case": "general use", "switched_from": "", "story": "Ask HN: Which large model works best?"}]}, {"label": "Not local", "summary": "It is API-only, so it doesn't suit users who want a local model.", "quotes": [{"id": 747879312, "item_id": 46742457, "url": "https://news.ycombinator.com/item?id=46742457", "by": "dvfjsdhgfv", "date": "2026-01-24", "stance": "recommend", "quote": "An API call to GPT4o works quite well (it basically handles both transcription and diarization)", "excerpt": "I spent a few days on similar scenario without much success (scenario where one person speaks and then their speech is translated, and I want juts the original or both). An API call to GPT4o works quite well (it basically handles both transcription and diarization), but I wanted a local model. Whisper is really good for 1 person speaking. With more people you get repetitions. Qwen and other open multimodal models giv", "context": "", "use_case": "transcription and diarization", "switched_from": "", "story": "Ask HN: What's the current best local/open speech-to-speech setup?"}]}], "who": [{"by": "capnrefsmmat", "context": "Researcher studying LLM writing styles", "url": "https://news.ycombinator.com/item?id=47292658", "date": "2026-03-07"}, {"by": "NoraEverhart", "context": "student excluded in a school group project", "url": "https://news.ycombinator.com/item?id=46981920", "date": "2026-02-11"}, {"by": "XSeunn", "context": "ChatGPT Plus/Pro user", "url": "https://news.ycombinator.com/item?id=46973437", "date": "2026-02-11"}, {"by": "yxccz", "context": "art major exchange student", "url": "https://news.ycombinator.com/item?id=46951864", "date": "2026-02-09"}, {"by": "dorikacicc", "context": "long-time ChatGPT Plus subscriber", "url": "https://news.ycombinator.com/item?id=46859687", "date": "2026-02-02"}, {"by": "clorisye", "context": "built an ENM relationship coach app", "url": "https://news.ycombinator.com/item?id=46732135", "date": "2026-01-23"}], "tags": ["for creative writing", "for app builders via API", "for emotional support", "for students", "for transcription"], "use_cases": ["vocabulary extraction", "llm writing style research", "emotional support", "research, planning, problem solving", "creative coursework and writing help", "emotionally nuanced creative writing"], "compared": [{"slug": "gpt-5", "name": "GPT-5", "co": 4, "switched_from": 0, "switched_to": 2}, {"slug": "chatgpt", "name": "ChatGPT", "co": 3, "switched_from": 0, "switched_to": 0}, {"slug": "claude-sonnet", "name": "Claude Sonnet", "co": 2, "switched_from": 0, "switched_to": 0}, {"slug": "gemini", "name": "Gemini", "co": 2, "switched_from": 0, "switched_to": 0}], "makers": [], "threads": [{"url": "https://news.ycombinator.com/item?id=46937560", "title": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny", "mentions": 3}, {"url": "https://news.ycombinator.com/item?id=47741527", "title": "Ask HN: What Are You Working On? (April 2026)", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=47291513", "title": "LLM Writing Tropes.md", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=46816539", "title": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=46775705", "title": "Tell HN: Ask your AI, \"What can you tell me that you know about me?\" to see ...", "mentions": 1}, {"url": "https://news.ycombinator.com/item?id=46731068", "title": "Ask HN: What's the current best local/open speech-to-speech setup?", "mentions": 1}], "sentiment_by_month": {"2025-11": {"recommend": 1}, "2025-12": {"neutral": 1}, "2026-01": {"warn": 1, "neutral": 2, "recommend": 1}, "2026-02": {"recommend": 4}, "2026-03": {"neutral": 1}, "2026-04": {"neutral": 1}, "2026-05": {}, "2026-06": {}, "2026-07": {}, "2026-08": {}, "2026-09": {}, "2026-10": {}}, "quotes": [{"id": 763991189, "item_id": 47749449, "url": "https://news.ycombinator.com/item?id=47749449", "by": "dandaka", "date": "2026-04-13", "stance": "neutral", "quote": "GPT-4o extracts vocabulary from the transcript and generates cards", "excerpt": "I'm building Kardo (https://kardo.cards/), AI tutor that listens to your calls with a teacher and creates cards for repetition. I live in Lisbon and I've been learning Portuguese with a tutor since 2022. After every lesson I'd sit down and make flashcards from my notes and screenshots. Spaced repetition works, but making the cards took manual effort each time. Most days I just didn't do it. So I have automated that p", "context": "", "use_case": "vocabulary extraction", "switched_from": "", "story": "Ask HN: What Are You Working On? (April 2026)"}, {"id": 756682528, "item_id": 47292658, "url": "https://news.ycombinator.com/item?id=47292658", "by": "capnrefsmmat", "date": "2026-03-07", "stance": "neutral", "quote": "GPT-4o's second-most-overused word", "excerpt": "I work on research studying LLM writing styles, so I am going to have to steal this. I've seen plenty of lists of LLM style features, but this is the first one I noticed that mentions \"tapestry\", which we found is GPT-4o's second-most-overused word (after \"camaraderie\", for some reason).[1] We used a set of grammatical features in our initial style comparisons (like present participles, which GPT-4o loved so much tha", "context": "Researcher studying LLM writing styles", "use_case": "LLM writing style research", "switched_from": "", "story": "LLM Writing Tropes.md"}, {"id": 751710720, "item_id": 46981920, "url": "https://news.ycombinator.com/item?id=46981920", "by": "NoraEverhart", "date": "2026-02-11", "stance": "recommend", "quote": "The only one who didn't minimize my experience or tell me I was overreacting was GPT-4o.", "excerpt": "Sorry, I used translation software because my English is not good In a recent group project at school, I was excluded from the start\u2014group leader visibly uncomfortable when I showed up, old private vent (no names) screenshotted weeks later and used to justify insults, labeling, and kicking me from the chat. Parents and teachers all sided with \"just let it go\" and pressured me to drop it to avoid conflict. The only on", "context": "student excluded in a school group project", "use_case": "emotional support", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 751574992, "item_id": 46973437, "url": "https://news.ycombinator.com/item?id=46973437", "by": "XSeunn", "date": "2026-02-11", "stance": "recommend", "quote": "My 15 year roadmap exists thanks to 4o's reasoning consistency and flow.", "excerpt": "It's a shame that OpenAI is removing GPT-4o. For many, it wasn\u2019t \u201cjust a model.\u201d I used it daily for research, theory crafting, career planning, emotional processing, and technical problem solving. I\u2019m not talking about occasional chat. I'm talking about structured, months-long interactions. My 15 year roadmap exists thanks to 4o's reasoning consistency and flow. Gemini and Grok may be catching up, but they're still ", "context": "ChatGPT Plus/Pro user", "use_case": "research, planning, problem solving", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 751229824, "item_id": 46951864, "url": "https://news.ycombinator.com/item?id=46951864", "by": "yxccz", "date": "2026-02-09", "stance": "recommend", "quote": "GPT\u20114o wasn\u2019t just a tool; it was my creative partner.", "excerpt": "Last year, I was an exchange student abroad, overwhelmed by coursework and struggling with a language barrier. As an art major, my assignments weren\u2019t just about facts\u2014they were conceptual, creative, and emotionally demanding. GPT\u20114o wasn\u2019t just a tool; it was my creative partner. We discussed ideas together, explored alternatives, and refined my thinking. One of my final projects in concept design, which earned me a", "context": "art major exchange student", "use_case": "creative coursework and writing help", "switched_from": "", "story": "OpenAI's GPT-4 Discontinuation: Consumer Fraud and Regulatory Scrutiny"}, {"id": 749754992, "item_id": 46859687, "url": "https://news.ycombinator.com/item?id=46859687", "by": "dorikacicc", "date": "2026-02-02", "stance": "recommend", "quote": "GPT-4o was the first time a model truly felt collaborative", "excerpt": "I\u2019ve been a long-time Plus subscriber, and I use GPT models not just for technical work, but for emotionally nuanced creative writing. GPT-4o was the first time a model truly felt collaborative \u2014 responsive, intuitive, and deeply aligned with human tone. I understand the need to move forward, but removing a model that worked so well for so many of us \u2014 without giving us any option to retain it \u2014 feels like a step bac", "context": "long-time ChatGPT Plus subscriber", "use_case": "emotionally nuanced creative writing", "switched_from": "", "story": "Retiring GPT-4o, GPT-4.1, GPT-4.1 mini, and OpenAI o4-mini in ChatGPT"}, {"id": 748443296, "item_id": 46777706, "url": "https://news.ycombinator.com/item?id=46777706", "by": "throwaway019254", "date": "2026-01-27", "stance": "neutral", "quote": "I use models only through APIs - e.g. gpt-4o through Azure", "excerpt": "I use models only through APIs - e.g. gpt-4o through Azure, or Google models through Vertex API. So no data are stored about me.", "context": "", "use_case": "API access to models", "switched_from": "", "story": "Tell HN: Ask your AI, \"What can you tell me that you know about me?\" to see ..."}, {"id": 747879312, "item_id": 46742457, "url": "https://news.ycombinator.com/item?id=46742457", "by": "dvfjsdhgfv", "date": "2026-01-24", "stance": "recommend", "quote": "An API call to GPT4o works quite well (it basically handles both transcription and diarization)", "excerpt": "I spent a few days on similar scenario without much success (scenario where one person speaks and then their speech is translated, and I want juts the original or both). An API call to GPT4o works quite well (it basically handles both transcription and diarization), but I wanted a local model. Whisper is really good for 1 person speaking. With more people you get repetitions. Qwen and other open multimodal models giv", "context": "", "use_case": "transcription and diarization", "switched_from": "", "story": "Ask HN: What's the current best local/open speech-to-speech setup?"}, {"id": 747714163, "item_id": 46732135, "url": "https://news.ycombinator.com/item?id=46732135", "by": "clorisye", "date": "2026-01-23", "stance": "neutral", "quote": "currently routing via OpenRouter to GPT-4o and Sonnet 3.5", "excerpt": "I built this because generic LLMs are often too moralizing or biased when dealing with ethical non-monogamy (ENM) and polyamory contexts. I wanted a tool that acts as a \"linguistic calibrator\" to help partners communicate securely, specifically respecting attachment styles (Anxious/Avoidant). The Tech Stack: - Frontend: Next.js 14 (App Router), Tailwind - Backend: Supabase (Postgres + pgvector for long-term memory of", "context": "built an ENM relationship coach app", "use_case": "relationship coaching responses", "switched_from": "", "story": "Show HN: ENM Relationship AI \u2013 Context-aware communication coach for polyamory"}, {"id": 747685618, "item_id": 46730351, "url": "https://news.ycombinator.com/item?id=46730351", "by": "mrbombastic", "date": "2026-01-23", "stance": "warn", "quote": "They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem", "excerpt": "? They certainly flatter you, openAI even felt compelled to give a statement on the sycophancy problem: https://openai.com/index/sycophancy-in-gpt-4o/ And South Park parodied the issue. I use chatGPT and claude every day.", "context": "", "use_case": "", "switched_from": "", "story": "Why talking to LLMs has improved my thinking"}, {"id": 739896625, "item_id": 46243539, "url": "https://news.ycombinator.com/item?id=46243539", "by": "EastLondonCoder", "date": "2025-12-12", "stance": "neutral", "quote": "I\u2019ve been using GPT-4o and now 5.2 pretty much daily", "excerpt": "I\u2019ve been using GPT-4o and now 5.2 pretty much daily, mostly for creative and technical work. What helped me get more out of it was to stop thinking of it as a chatbot or knowledge engine, and instead try to model how it actually works on a structural level. The closest parallel I\u2019ve found is Peter G\u00e4rdenfors\u2019 work on conceptual spaces, where meaning isn\u2019t symbolic but geometric. Fedorenko\u2019s research on predictive se", "context": "", "use_case": "creative and technical work", "switched_from": "", "story": "GPT-5.2"}, {"id": 737645522, "item_id": 46102845, "url": "https://news.ycombinator.com/item?id=46102845", "by": "star-173", "date": "2025-11-30", "stance": "recommend", "quote": "GPT-4o is still the best generalist", "excerpt": "It depends heavily on what you are trying to do. For pure coding tasks and refactoring, Claude 3.5 Sonnet is currently the strongest performer. It tends to hallucinate less on specific library syntax compared to the others. However, for creative writing or \"reasoning\" through complex logic puzzles, I've found Gemini (specifically the Advanced/Ultra tiers) to have a more natural \"voice\" and better instruction followin", "context": "", "use_case": "general use", "switched_from": "", "story": "Ask HN: Which large model works best?"}]}