Type: AI model
GPT-4 is OpenAI’s multimodal large language model introduced in 2023, a landmark release that set the standard for capable general-purpose AI and powered much of the early ChatGPT era. It remains a key reference point for measuring later model progress. GROUNDING tracks GPT-4 as a historical baseline against newer GPT-5-series models.
Recent Updates
- 2026-07-03: LearNAT proposes AST-guided decomposition for NL2SQL (cs.CL updates on arXiv.org) · arxiv.org
- 2026-07-03: Distributionally Robust Listwise Preference Optimization (cs.AI updates on arXiv.org) · arxiv.org — LLM Evals
- 2026-07-03: Prompt Engineering Guide adds web version, courses, and a larger resource index (GitHub AI Ranking Changes (Top 10)) · github.com — RAG Agents Context Engineering LLM Evals DAIR.AI DAIR.AI Academy · Google · OpenAI · ChatGPT Code Llama Flan · Gemini · LLaMA · Mistral-7B · Mixtral · Olmo Phi-2
- 2026-07-03: Four Years of Moving Goalposts for AI Coding Tools (Hacker News) · publicznyprofil.github.io — Code Agents LLM Evals Stack Overflow · OpenAI · Microsoft · Cognition · NVIDIA · Google · Alphabet Carl Brown Jensen Huang Sundar Pichai Andrej Karpathy Pieter Levels
- 2026-07-03: AI Productivity Gains Shrink in Real Payroll Data (Hacker News) · okaneland.com — LLM Evals BCG Harvard Anders Humlum Emilie Vestergaard · ChatGPT
- 2026-07-06: When AI Costs Outgrow Payroll (Hacker News) · tomtunguz.com — Agents Anthropic · OpenAI · Goldman Sachs · Ramp · Epoch AI KeyBanc Capital Markets OPEXEngine Levels.fyi U.S. Bureau of Labor Statistics · Fortune SaaStr · DeepSeek · Google · Claude · Gemini · DeepSeek-V3
- 2026-07-06: Quivr Enters GitHub Top 10: Opinionated Python RAG Framework with Multi-LLM Support (GitHub AI Ranking Changes (Top 10)) · github.com — RAG Reranking OpenAI · Anthropic · Mistral · Groq · Cohere Megaparse · Ollama · LLaMA · Gemma
- 2026-07-06: Price per 1M tokens is meaningless (Hacker News) · janilowski.pl — LLM Evals OpenAI · Anthropic · Artificial Analysis · GPT-4o · GPT 5.5 · Claude Opus 4.8 · Sonnet 5 · GLM-5.2 · DeepSeek V4 Pro · Fable 5
- 2026-07-07: Classification-head fine-tuning makes tiny Qwen3 models stronger on multiple-choice benchmarks (cs.LG updates on arXiv.org) · arxiv.org — arXiv · Qwen3 · Qwen3-0.6B · Qwen3-1.7B · GPT-3 PaLM
- 2026-07-07: Berkeley AI researchers argue near-free inference will reshape data systems around agents (The Berkeley Artificial Intelligence Research Blog) · bair.berkeley.edu — Agents Berkeley Artificial Intelligence Research Blog UC Berkeley EPIC Data Lab Aditya G. Parameswaran Abraham Lincoln
- 2026-07-09: PRD for AI features in 2026: what to specify beyond user stories (Все статьи подряд / Искусственный интеллект / Хабр) · habr.com — OpenAI · Anthropic gpt-4o-2024-08-06 · Claude claude-sonnet-4-6-20250214
- 2026-07-20: MAR-12 uses multi-angle reasoning to detect harmful humor in memes (cs.AI updates on arXiv.org) · arxiv.org — LLM Evals
- 2026-07-20: How Much It Really Costs to Train an LLM From Scratch (Все статьи подряд / Искусственный интеллект / Хабр) · habr.com — DeepSeek · DeepMind · Epoch AI · ByteDance · Habr Andrey Karpathy · DeepSeek-V3 · R1 habrGPT · GPT-2 PaLM · Chinchilla
- 2026-07-21: Scientific Feasibility Control for self-correcting scientific generation (cs.CL updates on arXiv.org) · arxiv.org — LLM Evals DeepSeek R1
FAQ
What is GPT-4?
GPT-4 is OpenAI’s multimodal large language model introduced in 2023, a landmark release that set the standard for capable general-purpose AI and powered much of the early ChatGPT era. It remains a key reference point for measuring later model progress. GROUNDING tracks GPT-4 as a historical baseline against newer GPT-5-series models.
What does this page track?
Dated radar mentions, source links, related concepts, and builder-relevant context for GPT-4, collected automatically by GROUNDING.
When was GPT-4 last mentioned?
GPT-4 was most recently mentioned in a radar update dated 2026-07-21.
Category: Text / Language Models