Blog

Hy4

❖ Hy4 preview is a new-generation Mixture-of-Experts (MoE) flagship model developed by the Tencent Hy Team. The model comprises 770B total parameters, of which 49B are activated per token. We scaled Hy4 on three fronts: model size, context length, and training data.

❖ GLM-5.3 is our latest flagship model for coding and long-horizon tasks. It marks a substantial leap in long-horizon task capability over its predecessor GLM-5.1 and delivers that capability on a solid 1M-token context. It is pure open with an MIT open-source license — no regional limits, technical access without borders.

Miles

❖ We present Miles v0.1, a full-stack production-ready system for frontier post-training, the successor to our first Miles release [1]. Building upon slime’s clean design, Miles optimizes every stage in the RL training loop around a simple principle: verified, clean, and customizable everywhere. With accuracy, efficiency, reliability, and scalability as first-class goals, Miles aims to make frontier-scale RL accessible to researchers and developers alike. In this blog post, we will walk through Miles end-to-end.

Deep Researcher

❖ The NVIDIA Deep Researcher Agent Blueprint is a deployable research backend built on the NVIDIA NeMo Agent Toolkit and LangChain Deep Agents. Teams can self-host the application boundary and connect deployment-owned models, data sources, authentication, policy controls, storage, and observability. It provides both quick, cited answers and in-depth, report-style research, plus benchmarks and evaluation harnesses for measuring quality.

Speculators

❖ Speculators is a library for training speculative decoding draft models that deploy directly to LLM inference engines like vLLM. Speculative decoding is a lossless technique that speeds up LLM inference by using a smaller, faster draft model (“the speculator”) to propose tokens, which are then verified by the larger base model, reducing latency without compromising output quality.

SpecForge

❖ Today, we are introducing a major update to SpecForge. The new runtime separates target-model inference from draft-model training, supports a broader family of speculative decoding algorithms, and unifies online, offline, and disaggregated workflows behind one typed training entry point. Alongside the release, we are publishing more draft models, covering different speculative decoding methods and different target models.

Qwen3.8

❖ We are officially releasing Qwen 3.8-Max, the most capable model in the Qwen family to date. This also marks the first time we will open-source the weights of a Qwen-Max-class model. Built upon the architectural foundation of Qwen 3.5, Qwen 3.8-Max scales to 2.4 trillion parameters, delivering comprehensive improvements across coding, work, research, and long-horizon tasks.

❖ We introduce Kimi K3 — our most capable model. Kimi K3 is a 2.8T-parameter model built on our Kimi Delta Attention and Attention Residuals, with native vision capabilities and a 1-million-token context window. It is the world’s first open 3T-class model, designed for frontier intelligence across long-horizon coding, knowledge work, and reasoning.