#coding agents
Coding Agents: 24 AI articles covering coding agents news, analysis, and research
Articles
Microsoft Open-Sources Agensh: A Decentralized Multi-Agentβ8
Microsoft Research open-sources Agensh, a decentralized multi-agent harness with no central orchestrator that scales to 1,024 coding agents and boosts hard Prog...
Vercel Open-Sources durable-harness for Crash-Proof Coding Agentsβ9
Vercel open-sources durable-harness, a package that keeps coding agent sessions alive through crashes and timeouts by logging each session durably.
StepFun Releases Step 5 Preview as Open Weights: 600B MoE, 27Bβ9
StepFun releases Step 5 Preview open weights: a 600B MoE model with 27B active parameters, 1M-token context, and strong long-context reasoning for agentic workl...
Alibaba Qwen Introduces QwenGyre: Elastic Reinforcement Learningβ9
Alibaba's Qwen team introduces QwenGyre, an elastic reinforcement learning framework for extreme-long-horizon agents, delivering up to 1.85x speedup on long cod...
The State of AI and Software Engineering in 2026: Code Is Cheap,β10
AI has cut code generation costs dramatically, but a 2026 report shows human expertise, judgment, and accountability remain the real bottleneck in software engi...
Together Link: A Free CLI That Runs Open Models Like Kimi K3 andβ9
Together Link is a free MIT-licensed CLI that connects Claude Code, Codex, and OpenCode to open models like Kimi K3 and GLM 5.3 hosted on Together AI.
NVIDIA's SoL-Pi Auto-Research Loops Cut Coding Agent Tokenβ9
NVIDIA's SoL-Pi auto-research loops cut coding agent token traffic by up to 49% on EdgeBench, reducing API costs ~33% while matching Pi's scores.
How to Maximize GPT-6 Astraβ8
Hands-on GPT-6 Astra tips: maximize coding, research, and automation, plus first impressions, drawbacks, and workflow integration strategies.
How to Run 10+ Claude Code Sessions Without a Powerful Computerβ10
Run 10+ Claude Code sessions without a powerful PC by offloading agents to cloud servers. Save local resources, enable 24/7 uptime, and cut hardware costs.
Give Your Coding Agents a Memory You Ownβ9
Give your coding agents persistent memory you controlβboost efficiency, preserve project context across sessions, and overcome stateless AI limitations.