2026
7 篇文章把銀行 API 上線流程拆成七個專責代理：Ninth Wave 在 Amazon Bedrock 上的多代理設計
Ninth Wave 用 Amazon Bedrock AgentCore 把開放金融 onboarding 拆成七個專責代理，以租戶隔離與確定性評分解決合規痛點。
閱讀文章 ↗AMIE 的下一步：AI 醫療從文字走向即時視訊問診
Google Research 與 Google DeepMind 在 2026 年 8 月發表 AMIE 的視訊問診研究，展示 AI 如何透過多代理架構理解視覺與聽覺線索。本文從產品建造者角度，拆解這項技術對醫療 AI 產品設計的啟示與限制。
閱讀文章 ↗Codex 5.6 簡評：模型分工、Sol Ultra 與更快見底的額度
從 GPT-5.6 Sol、Terra、Luna 的定位，到 Sol Ultra multi-agent 工作流的實測用量：價格相同，不代表消耗相同。
閱讀文章 ↗A2A 協定滿一週年:150 家組織、五種 SDK 與 v1.0 穩定規格
2026 年 4 月 9 日 A2A 協定滿一週年:Linux Foundation 宣布支持組織從 50+ 增至 150+,SDK 擴至五種語言,v1.0 加入 Signed Agent Cards,並已落地 Azure AI Foundry 與 Bedrock AgentCore。
閱讀文章 ↗Grok 4.20 公測：四個代理人先辯論再回話
xAI 於 2026 年 2 月 17 日推出 Grok 4.20 公開測試：內建 Grok、Harper、Benjamin、Lucas 四個代理人，推理時即時辯論、交叉驗證後才輸出。eWeek 報導幻覺下降約 65%，本文拆解運作機制、成本與對開發者的意義。
閱讀文章 ↗16 個 Claude 代理寫出 C 編譯器:兩週、10 萬行、2 萬美元
Anthropic 研究員 Nicholas Carlini 以 16 個平行 Claude Opus 4.6 代理、近 2,000 個工作階段與約 2 萬美元,在兩週內寫出 10 萬行 Rust 的 C 編譯器,可編譯出可開機的 Linux 6.9 核心。本文拆解其極簡編排與限制。
閱讀文章 ↗Kimi K2.5 開源釋出：原生視覺加上 Agent Swarm 多代理協作
Moonshot AI 於 2026 年 1 月 27 日發表 Kimi K2.5：MIT 授權的開源 MoE model，具備原生視覺與 Agent Swarm 多代理協作。本文解析它對自架多模態 agent 堆疊與多代理系統生態的意義。
閱讀文章 ↗
2026
7 ARTICLESWhat a Multi-Agent Onboarding Assistant Changes for Open Finance Integration
Ninth Wave's Compass uses per-task models and tenant-scoped grounding to cut API onboarding from weeks to a self-service workflow.
READ POST ↗AMIE: Google's Medical AI Moves from Text to Real-Time Video Consultations
Google Research and DeepMind advance AMIE to real-time video consultations, using multi-agent architecture to interpret visual and auditory cues in simulated clinical settings.
READ POST ↗Codex 5.6 Review: Model Tiers, Sol Ultra, and Faster Limit Burn
A practical comparison of GPT-5.6 Sol, Terra, and Luna—and why Sol Ultra multi-agent work can exhaust a Codex allowance much faster even when the headline rate matches GPT-5.5.
READ POST ↗A2A Turns One: 150+ Orgs, Five SDKs, Stable v1.0 Spec
A2A turned one on April 9, 2026: 150+ orgs, five production SDKs, a stable v1.0 spec with Signed Agent Cards, and deployments on Azure AI Foundry and Amazon Bedrock AgentCore.
READ POST ↗Grok 4.20 Beta: Four Agents That Debate Before They Answer
xAI's Grok 4.20 beta ships four named agents that debate in real time before answering. eWeek reports hallucinations down about 65%. Here is how the loop works and what it costs builders.
READ POST ↗16 Claude Agents Built a C Compiler: 100K Lines, $20K
Anthropic researcher Nicholas Carlini ran 16 parallel Claude Opus 4.6 agents over ~2,000 Claude Code sessions to write a 100,000-line Rust C compiler that boots Linux 6.9 — for under $20,000.
READ POST ↗Kimi K2.5 Goes Open Source: Native Vision and Agent Swarm Coordination
Moonshot AI released Kimi K2.5 on January 27, 2026: an MIT-licensed open-source MoE model with native vision and Agent Swarm coordination. What it means for self-hosted agent stacks.
READ POST ↗