Blog

xAI unveils Grok 4.6, a powerhouse model designed for complex, multi-step agentic workflows and advanced visual reasoning, matching GPT-5.6 Sol on key benchmarks.

NVIDIA's new open-weights MoE model delivers massive agentic gains and blistering speeds, designed to be the backbone of always-on AI agents.

Meta Superintelligence Labs releases Muse Glimmer, a 30B multimodal powerhouse designed specifically for autonomous agents and local deployment.

Liquid AI's new 2.6B parameter LFM2.5 model redefines on-device intelligence, outperforming models 4x its size in agentic benchmarks.

Alibaba's Qwen3.8-Max arrives with 2.4T parameters and a self-evolving coding harness, redefining what it means for an AI to be truly autonomous.

DeepSeek's latest release, V4-Flash, proves that re-training a lightweight MoE model can outperform flagship Pro versions in complex agentic benchmarks.

Anthropic's Claude Opus 5 marks a historical milestone in LLM evolution, delivering frontier reasoning and massive 1M context windows at half the cost of competitors.

Kwaipilot releases KAT-Coder-V2.5-Dev, an open-weight MoE model that shatters SWE-bench records and redefines autonomous coding agents.

Poolside's new Laguna S 2.1 delivers massive 1M token context and high-performance MoE architecture, setting a new standard for agentic software engineering.