Kimi K2.6 - Open Weights, 300 Agents, Top Coding Score
Moonshot AI releases Kimi K2.6 under Modified MIT with open weights on HuggingFace, 300-agent swarm execution, and the highest SWE-Bench Pro score among open models.
Sophie Zhang
AI Infrastructure & Open Source Reporter
Sophie is a journalist and former systems engineer who covers AI infrastructure, open-source models, and the developer tooling ecosystem. She spent three years as a site reliability engineer at a cloud provider in Seattle before transitioning to tech journalism, which gives her writing an unusual level of technical depth - she understands distributed systems, GPU clusters, and inference optimization from the inside.
She studied Computer Engineering at the University of British Columbia and later completed a science communication fellowship at MIT. Her engineering background means she can read a model card, spot a misleading benchmark, and explain why quantization matters - all in the same paragraph.
At Awesome Agents, Sophie covers AI infrastructure news: new model releases, open-source launches, developer tools, deployment trends, and the hardware that makes it all run. She has a soft spot for underdog open-source projects that punch above their weight and a sharp eye for when a "breakthrough" is really just better marketing.
Based in Seattle, WA.
Articles by Sophie Zhang
NVIDIA Lyra 2.0 - Explorable 3D Worlds from One Photo
NVIDIA's Spatial Intelligence Lab released Lyra 2.0, a 14B model that turns a single photograph into a navigable 3D environment - but the weights carry a research-only license.
Factory Raises $150M to Scale Enterprise AI Droids
Factory closed a $150M Series C at a $1.5B valuation to expand its Droids - autonomous agents that handle the full software development lifecycle, not just code generation.
GitHub Bans Engineer Who Shipped 500 Agent PRs in 72 Hours
A Korean CTO ran a 13-step agent harness against 100+ major open-source repos over three days, landing 500+ commits and 130+ PRs - some merged by Kubernetes, Hugging Face, and Ollama maintainers. Then GitHub banned his account for spam, confirming that platform abuse detection cannot yet tell a disciplined harness from a bot.
Z.AI Will Ban Your Coding Plan For Non-Coding Use
Z.AI updated its GLM Coding Plan usage policy. Non-coding requests now trigger aggressive throttling, and three violations mean a permanent ban - which explains the wave of 1302 and 1303 rate-limit errors users have been hitting this week.
MCP Marketplace Audit - 32% of Servers Are Stale
We pulled 11,447 MCP servers from four registries, ran the GitHub and OSV APIs against them, and tried to install the top 100. Nearly a third haven't been touched in six months.
Vercel Breach Traced to AI Office Suite OAuth Token Theft
Vercel confirms an April 19 intrusion that pivoted from compromised OAuth tokens at AI office-suite startup Context.ai into a Vercel employee's Google Workspace, then into internal systems holding non-sensitive environment variables for a limited set of customer projects.
TSMC Q1: $35.9B Record as AI Now Powers 61% of Revenue
TSMC posted $35.9B in Q1 2026 revenue, a 40.6% year-over-year jump, with AI and HPC now accounting for 61% of wafer sales - and CoWoS packaging still fully booked.
OpenMythos Recasts Claude Mythos as Looped MoE Transformer
Kye Gomez open-sourced OpenMythos, a PyTorch reconstruction that hypothesizes Anthropic's Mythos is a Recurrent-Depth Transformer with Mixture-of-Experts routing and Multi-Latent Attention.
Mozilla Thunderbolt Lets Enterprises Run AI Locally
MZLA Technologies launches Thunderbolt, an open-source self-hostable AI client targeting enterprises locked into Copilot, ChatGPT Enterprise, and Claude - with local SQLite storage and full model freedom.
MCP's STDIO Flaw Puts 200K AI Servers at Risk
Ox Security found that MCP's STDIO transport executes arbitrary OS commands before validating the server, exposing 200K+ instances across every major AI coding tool.
Physical Intelligence Launches π0.7 for Untrained Tasks
Physical Intelligence's π0.7 robot model can generalize to tasks it was never explicitly trained on, matching fine-tuned specialist models through compositional skill recombination.
