We built a cyber-capable AI model that, based on benchmarking and real-world testing, is perhaps the most capable offensive-security LLM we have seen It scores 95% on Cybench, solving 38/40 tasks But benchmarks only go so far, so we pointed it at real systems Through authorized programs, we tested it against a Fortune-500 technology company, a major global telecom, and leading Web3 protocols. The result: a 100% success rate - it found a viable path into every authorized target in our test set. It adapts, reasons, and finds routes that aren’t obvious, even in systems that already passed audits and security reviews. If this holds at scale, most current security assumptions break.

Introducing Kimi K3: Open Frontier Intelligence 🔹 2.8 Trillion Parameters, 1 Million Context, Native Multimodal 🔹 Kimi Delta Attention enables up to 6.3x faster decoding in million-token contexts 🔹 Attention Residuals deliver ~25% higher training efficiency at <2% additional cost 🔹 Built for long-horizon agentic coding and self-evolving workflows Kimi K3 is now live on on Kimi.com, Kimi Work, Kimi Code, and the Kimi API. Open Weights by July 27, 2026. 🔗 API: platform.kimi.ai 🔗 Tech blog: kimi.com/blog/kimi-k3

Introducing GLM-5.2: Frontier Intelligence, Open Weights - Significant improvements in coding and agentic tasks - Strong long-horizon capabilities with a 1M context window - Two levels of reasoning effort: GLM-5.2 (max) pushes the limits, while GLM-5.2 (high) strikes a strong balance between performance and token efficiency - MIT-licensed open weights - Same API pricing as GLM-5.1 Tech Blog: z.ai/blog/glm-5.2 Weights: huggingface.co/zai-org/GLM-5… API: docs.z.ai/guides/llm/glm-5.2 Coding Plan: z.ai/subscribe Chat: chat.z.ai