Hackernews posts about GPT-4

GPT-4 is an advanced artificial intelligence model developed by OpenAI that enables natural language processing and text generation capabilities.

1. From GPT-4 to GPT-5: Measuring progress through MedHELM [pdf] (www.fertrevino.com)
   127 points by fertrevino 27 days ago | 96 comments
2. Show HN: I fine-tuned GPT4.1 on my iMessage history (jonyork.net)
   10 points by jonpizza 21 days ago | 6 comments
3. Study: GPT-4V can evaluate social situations similar to humans (www.utu.fi)
   5 points by giuliomagnifico 8 days ago | 1 comments
4. Open‑source InternVL3.5 crushes GPT‑4V on multimodal benchmarks (medium.com)
   4 points by acossta 22 days ago | 1 comments
5. Ask HN: Why is GPT-5 almost the same as GPT-4?
   3 points by whyandgrowth about 14 hours ago | discuss
6. Nano Banana vs. GPT-4o: side-by-side AI image comparisons (github.com)
   2 points by mustaphah 10 days ago | discuss
7. MiniCPM-V 4.5: GPT-4o Level MLLM for Image and Video Understanding on Your Phone (huggingface.co)
   2 points by amram_art 22 days ago | discuss
8. Website lets you blind-test GPT-5 vs. GPT-4o (venturebeat.com)
   1 points by isaacfrond 23 days ago | 2 comments
9. I still use GPT 4o (firasd.substack.com)
   1 points by firasd 8 days ago | discuss
10. Notes on Image Generation with GPT-4.1 (taoofmac.com)
    1 points by rcarmo 28 days ago | discuss
11. Launch HN: Dedalus Labs (YC S25) – Vercel for Agents
    77 points by windsor 20 days ago | 15 comments
12. Show HN: TheAuditor – Offline security scanner for AI-generated code (github.com)
    13 points by TheAuditorTool 10 days ago | 32 comments
13. Prompt engineering is collapsing – GPT-5 just proved it
    10 points by yuer2025 24 days ago | 8 comments
14. Show HN: Pluely – open-source alternative to Cluely (~10MB, Always On Display) (github.com)
    7 points by truly_sn 30 days ago | 5 comments
15. Show HN: Testronaut – AI-powered mission-based browser testing (testronaut.app)
    4 points by scfast 21 days ago | 1 comments
16. Show HN: Dr. WFGY – an AI doctor for broken pipelines (chat-based） (chatgpt.com)
    3 points by tgrrr9111 11 days ago | 2 comments
17. High rate of LLM (GPT5) hallucinations in dense stats domains (cricket)
    3 points by sp1982 22 days ago | 2 comments
18. Show HN: I used n-grams to see which LLM writes most like a human (kite-mu.vercel.app)
    3 points by shadowblue 1 day ago | discuss
19. Modular LLM framework inspired by Linux – aiming for a one-GPU future
    2 points by openkame 22 days ago | 1 comments
20. Show HN: I built an open-source CSV importer that I wish existed (github.com)
    2 points by aray07 18 days ago | discuss
21. Show HN: Llmswap v3.0 – CLI and SDK for OpenAI, Claude, Gemini, Watsonx (pypi.org)
    2 points by sreenathmenon 28 days ago | discuss
22. GPT5 is the best coding LLM because other LLMs admit it?
    1 points by adinhitlore 22 days ago | 8 comments
23. GPT-5 Doesn't Vibe?
    1 points by anupshinde 17 days ago | 3 comments
24. Ask HN: What in your opinion is the best model for vibecoding? My thoughts below
    1 points by adinhitlore 1 day ago | 2 comments
25. Ask HN: Are AI development costs pushing you to seek alternatives?
    1 points by uto_31 20 days ago | 2 comments
26. Show HN: Theory of Mind benchmark for 8 LLMs with reproducible markers
    1 points by AlekseN 7 days ago | 1 comments
27. $1T Agent Interoperability in Plain Sight
    1 points by Antonioromero10 29 days ago | 1 comments
28. Show HN: Cut AI API costs 90% with intelligent model routing (apicrusher.com)
    1 points by bytecounter 2 days ago | discuss
29. Show HN: DailyMastery – AI-powered personalized daily learning lessons (dailymastery.io)
    1 points by dalenguyen 7 days ago | discuss
30. Show HN: Turn Your LinkedIn into a Professional Website in 60 Seconds (www.craftwebbb.com)
    1 points by FlorinDobinciuc 13 days ago | discuss