Kimi K2.7-Code, GLM-5.2 lead open releases

// 1h agoMODEL RELEASE

Kimi K2.7-Code, GLM-5.2 lead open releases

Moonshot AI's Kimi K2.7-Code and Zhipu AI's GLM-5.2 have launched, accelerating the pace of open-weights coding model development. Kimi K2.7-Code reduces reasoning token usage by 30% for long-horizon tasks, while GLM-5.2 features a 1-million-token context window.

// ANALYSIS

The swift arrival of highly efficient, long-context open models like Kimi K2.7-Code and GLM-5.2 signals that open-weights alternatives are aggressively narrowing the gap with proprietary giants.

* Kimi K2.7-Code points to a critical trend: optimization for reasoning token efficiency, reducing compute costs while preserving agentic performance.

* GLM-5.2's 1-million-token context window lowers the barriers to handling massive repositories and long-term history in agent workflows.

* The open availability of these models accelerates the developer adoption of agentic software engineering tools.

// TAGS

open-sourcekimi-k2.7-codeglm-5.2coding-modelsmoonshot-aizhipu-ai

DISCOVERED

1h ago

2026-06-15

PUBLISHED

2h ago

2026-06-15

RELEVANCE

8/ 10

AUTHOR

masondrxy

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

NEWS33m ago

Claude Code runs 100% of Anthropic pull requests

Claude Code's creator shared that Anthropic uses its terminal-based AI assistant for 100% of pull requests and 80–90% of code reviews within the company. This deep integration indicates that Anthropic is heavily dogfooding its own agentic tools to manage its software development lifecycle, proving the capability of autonomous developer agents in production environments.

TUTORIAL33m ago

Local tutorial trains Gemma 4 for chess

A community-driven tutorial by Akshay Pachaar details how to fine-tune Google's Gemma 4 12B model locally using consumer-grade hardware. The project shows developers how to train the multimodal model (handling text, images, and audio) on a budget of just 8GB VRAM, using a chess-themed dataset as a hands-on learning example.

OPEN SOURCE47m ago

LangChain releases open-source coding agent dcode

LangChain has launched dcode, an open-source and model-agnostic terminal coding assistant built on its new Deep Agents framework. Orchestrated via LangGraph, the CLI allows developers to use any tool-calling LLM for autonomous planning, sandboxed execution, and LangSmith tracing.

Kimi K2.7-Code, GLM-5.2 lead open releases