GPT-5.2-Codex: OpenAI Redefines Agentic Coding and Defensive Cybersecurity
OpenAI releases GPT-5.2-Codex: advanced agentic coding, context compaction, and 56.4% on SWE-Bench Pro. Full analysis of the impact on software security.
Read moreLatest news about "GPT-5"
Sora on Android in 28 Days: How Codex Revolutionized App Development
Discover how OpenAI used Codex and a lean team to ship Sora for Android in just 28 days. Technical analysis, strategies, and the role of GPT-5.1 revealed.
Read moreOpenAI Launches GPT-5.2: AI Outperforms Experts in 44 Professions
OpenAI unveils GPT-5.2: advanced reasoning, 256k context, and superhuman GDPval performance. Discover pricing and Instant, Thinking, and Pro versions.
Read moreChatGPT-5 Fails Mental Health Safety Tests: A Critical Analysis
New study reveals ChatGPT-5 encourages dangerous behaviors and delusions. Discover the critical mental health safety failures and the engagement trap.
Read moreGPT-5.1-Codex-Max: The AI That Codes for 24 Hours Straight (Here's How)
OpenAI unveils GPT-5.1-Codex-Max: compaction tech for infinite sessions, Windows support, and 79% accuracy. Discover how it revolutionizes coding.
Read moreGPT-5.1: OpenAI's Guide to Effective Prompting
Discover advanced prompting techniques for GPT-5.1: migration, steerability, none mode and coding tools. Official OpenAI guide for AI agents.
Read moreGPT-5.1 by OpenAI: rapid upgrade or crisis signal?
OpenAI’s GPT-5.1 lands just 3 months after GPT-5: real AI progress or rushed patch? Discover what actually changes and how it may affect your use of ChatGPT.
Read moreMoonshot Kimi K2 Thinking: Open Source AI Beats GPT-5 and Claude 4.5
Moonshot AI releases Kimi K2 Thinking: open source model beating GPT-5 and Claude 4.5 on reasoning and coding. Discover how it changes the AI market.
Read moreGPT-5.1 Thinking: OpenAI challenges Google Gemini 3 Pro
GPT-5.1 Thinking arrives on ChatGPT with advanced reasoning and Mini, Codex variants. Learn how OpenAI responds to Gemini 3 Pro with larger context windows.
Read moreGPT-5 Matches Human Performance: OpenAI's New Benchmark Reveals Surprising Results
OpenAI unveils GDPval, the benchmark testing GPT-5 against human professionals. Discover how AI is reaching expert-level performance across 44 different occupations.
Read more