Engineering Radar

#llm

3 signals tagged llm

AI & LLMs

HighEvaluate

The Model Context Protocol drops its stateful past for a stateless core

MCP's 2026-07-28 spec turns a bidirectional, session-based protocol into a stateless request/response one — remote MCP servers can now run behind a plain load balancer.

Jul 28StaffPrincipalArchitect

Source ↗

AI & LLMs

HighEvaluate

Kimi K3 puts a 1M-token context window into open weights

Moonshot AI released the open weights of Kimi K3 — a 2.8-trillion-parameter multimodal MoE with a 1,048,576-token context window — bringing frontier-scale long context to a model you can self-host.

Jul 27StaffPrincipalArchitect

Source ↗

Research → Prod

HighLearn now

Reasoning-by-RL goes peer-reviewed: DeepSeek-R1 in Nature

The method behind today's reasoning models — incentivizing step-by-step reasoning through pure reinforcement learning, with no human reasoning traces — was published, peer-reviewed, in Nature.

Sep 18, 2025SeniorStaffPrincipalArchitect

Source ↗

All signals