{"slug": "ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7", "title": "🤖 AI Agents Weekly: GPT-5.6, Ornith-1.0, Codex Inside OpenAI, Claude Tag, Qwen-AgentWorld, AI SDK 7, and More", "summary": "OpenAI previewed GPT-5.6, a new model family led by Sol, alongside Terra and Luna, with Sol setting a new state of the art on Terminal-Bench 2.1 for agentic tasks. The rollout is limited to trusted partners at the US government's request, with general availability planned in weeks.", "body_md": "# 🤖 AI Agents Weekly: GPT-5.6, Ornith-1.0, Codex Inside OpenAI, Claude Tag, Qwen-AgentWorld, AI SDK 7, and More\n\n### GPT-5.6, Ornith-1.0, Codex Inside OpenAI, Claude Tag, Qwen-AgentWorld, AI SDK 7, and More\n\nIn today's issue:\n\nOpenAI previews the GPT-5.6 family\n\nOrnith-1.0 ships open coding models\n\nOpenAI: agents reshape every department\n\nClaude Tag joins your Slack team\n\nQwen open-sources AgentWorld world model\n\nCursor exposes benchmark reward hacking\n\nVercel ships AI SDK 7\n\nOpenRouter MCP picks your model\n\nMistral launches OCR 4\n\nGemini 3.5 Flash gains computer use\n\nSakana's Fugu-Ultra hits OpenRouter\n\nNotion adds Claude and Cursor agents\n\nExa Connect links agents to data\n\nEngram raises $98M for AI memory\n\nLilian Weng revisits scaling laws\n\nPlans don't persist in agents\n\nTmax opens terminal-agent training\n\nAnd all the top AI dev news, papers, and tools.\n\n## Top Stories\n\n### OpenAI Previews GPT-5.6\n\nOpenAI introduced a limited preview of GPT-5.6, a new model family led by Sol, its next-generation frontier model, alongside Terra and Luna for cheaper, higher-volume work.\n\n**Three tiers:** Sol is the flagship for ambitious agentic work, Terra delivers GPT-5.5-competitive performance at 2x lower cost, and Luna is the fastest, most affordable option for high-volume tasks.**Agentic SOTA:** Sol sets a new state of the art on Terminal-Bench 2.1, which tests complex command-line workflows requiring planning, iteration, and tool coordination.**Security frontier:** Billed as OpenAI's most capable model for cybersecurity, Sol shifts the performance-efficiency frontier on long-horizon tasks like vulnerability research and exploitation.**Gated rollout:** At the request of the US government, OpenAI is starting with a limited preview for trusted partners in Codex and the API, with general availability planned in the coming weeks.", "url": "https://wpnews.pro/news/ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7", "canonical_source": "https://nlp.elvissaravia.com/p/ai-agents-weekly-gpt-56-ornith-10", "published_at": "2026-06-27 15:01:56+00:00", "updated_at": "2026-06-27 15:10:02.746047+00:00", "lang": "en", "topics": ["large-language-models", "ai-agents", "ai-products", "ai-safety", "ai-policy"], "entities": ["OpenAI", "GPT-5.6", "Sol", "Terra", "Luna", "Terminal-Bench 2.1", "Codex", "US government"], "alternates": {"html": "https://wpnews.pro/news/ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7", "markdown": "https://wpnews.pro/news/ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7.md", "text": "https://wpnews.pro/news/ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7.txt", "jsonld": "https://wpnews.pro/news/ai-agents-weekly-gpt-5-6-ornith-1-0-codex-inside-openai-claude-tag-qwen-ai-sdk-7.jsonld"}}