{"slug": "ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small", "title": "🤖 AI Agents Weekly: Kimi K3, DeepSeek-V4-Flash API, GPT-5.6 Price Cuts, Inkling-Small, YC's QM Harness, Gemini Robotics 2, Codex Security CLI, and More", "summary": "Moonshot AI released Kimi K3, a 2.8T-parameter open-weight mixture-of-experts model with native vision, a 1-million-token context window, and 104B active parameters per token, achieving frontier-level performance on long-horizon coding, agentic, reasoning, and vision tasks, trailing only Claude Fable 5 and GPT-5.6 Sol among models evaluated. The model uses Kimi Delta Attention, Attention Residuals, and Stable LatentMoE, with million-token agentic reinforcement learning and multiple reasoning-effort levels.", "body_md": "# 🤖 AI Agents Weekly: Kimi K3, DeepSeek-V4-Flash API, GPT-5.6 Price Cuts, Inkling-Small, YC's QM Harness, Gemini Robotics 2, Codex Security CLI, and More\n\n### Kimi K3, DeepSeek-V4-Flash API, GPT-5.6 Price Cuts, Inkling-Small, YC's QM Harness, Gemini Robotics 2, Codex Security CLI, and More\n\nIn today's issue:\n\nMoonshot open-sources Kimi K3\n\nDeepSeek ships V4-Flash agent API\n\nOpenAI cuts GPT-5.6 prices 80%\n\nThinking Machines drops Inkling-Small\n\nGoogle launches Gemini Robotics 2\n\nOpenAI open-sources Codex Security CLI\n\nYC open-sources its QM agent harness\n\nMicrosoft Foundry adds tool search\n\nMoonshot releases agent RL infra\n\nNous adds wake word to Hermes\n\nCursor lands on iPad\n\nResearchArena probes AI R&D sabotage\n\nHANDBOOK.md tests long policy files\n\nStudy exposes coding agent harness effects\n\nSlopCodeBench stress-tests Opus 5\n\nAnd all the top AI dev news, papers, and tools.\n\n## Top Stories\n\n### Moonshot Open-Sources Kimi K3\n\nMoonshot AI released Kimi K3, a 2.8T-parameter open-weight MoE model with native vision that lands closer to the closed frontier than any prior open release.\n\n**Architecture:** Combines Kimi Delta Attention, Attention Residuals, and Stable LatentMoE, activating 16 of 896 routed experts and 104B parameters per token.**Scale and context:** Ships a 1-million-token context window and roughly 2.5x better scaling efficiency than Kimi K2.**Agentic post-training:** Uses million-token agentic RL with persistent rollout and sandbox state, plus multiple reasoning-effort levels for long-horizon execution.**Where it lands:** Frontier-level on long-horizon coding, agentic, reasoning, and vision tasks, trailing only Claude Fable 5 and GPT-5.6 Sol among models evaluated.", "url": "https://wpnews.pro/news/ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small", "canonical_source": "https://nlp.elvissaravia.com/p/ai-agents-weekly-kimi-k3-deepseek", "published_at": "2026-08-01 15:02:40+00:00", "updated_at": "2026-08-01 15:36:59.369634+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-research", "ai-products"], "entities": ["Moonshot AI", "Kimi K3", "Claude Fable 5", "GPT-5.6 Sol"], "alternates": {"html": "https://wpnews.pro/news/ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small", "markdown": "https://wpnews.pro/news/ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small.md", "text": "https://wpnews.pro/news/ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small.txt", "jsonld": "https://wpnews.pro/news/ai-agents-weekly-kimi-k3-deepseek-v4-flash-api-gpt-5-6-price-cuts-inkling-small.jsonld"}}