Blog

Tencent has officially released Hy3, a massive 295B MoE model that brings frontier-level performance to the open-source community under Apache 2.0.

With a massive jump in SWE-bench scores and a fully permissive license, Laguna XS 2.1 is redefining what small-scale Mixture-of-Experts models can achieve in software engineering.

Meituan's LongCat-2.0 breaks the ceiling for open-source coding models, featuring a 1.6T MoE architecture and outperforming GPT-5.5 on SWE-bench Pro.

Liquid AI's new 230M parameter LFM2.5-230M is redefining on-device AI, delivering high-speed agentic performance on everything from smartphones to robotics.

Deep Reinforce AI unveils Ornith-1.0, a milestone family of agentic coding models achieving state-of-the-art performance across four parameter configurations, matching Claude Opus 4.7 while remaining fully open-source under MIT license.

Sakana AI's Fugu system introduces a novel approach to AI by orchestrating multiple language models through a single API, achieving frontier-level performance while providing vendor independence.

Zhipu AI's GLM-5.2, released on 2026-06-16, is a MIT-licensed flagship language model with a 1M-token context, 128K-token output, and IndexShare architecture. For developers, it signals a new era of open-source, project-scale coding agents and long-context reasoning systems.

Moonshot AI’s kimi-k2.7-code brings open-weight coding, 256k context, multimodal input, tool calling, JSON mode, partial mode, and cheaper cached inference for developers building code agents.

Cohere's new North Mini Code model brings elite reasoning and a massive 256K context window to the open-source coding ecosystem.