# 🤖 AI Agents Weekly: OpenAI Agents API, Cognition SWE-2, DeepSeek-V4.1-Flash, Cursor Projects, Sakana Fugu Max, Meta Muse, and More

> Source: <https://nlp.elvissaravia.com/p/ai-agents-weekly-openai-agents-api>
> Published: 2026-09-12 15:44:32+00:00

In today’s issue:

- OpenAI launches the Agents API
- Cognition ships SWE-2
- DeepSeek drops V4.1-Flash
- Sakana ships Fugu Max and Ultra v2
- Meta launches Muse personal agent
- GPT-Live-1 arrives in the API
- Claude Code adds plugin evals
- ChatGPT Work gets a Data agent
- Cursor introduces Projects
- AutoResearchExam tests 24-hour research agents

And all the top AI dev news, papers, and tools.

## Top Stories

### OpenAI Agents API

OpenAI released the Agents API in public beta, giving developers the same harness and infrastructure that run Codex. OpenAI operates the agent loop (model calls, tool use, and context), and developers pick the tools and where code runs.

- **Managed Codex harness:** The harness is versioned alongside model launches and handles context compaction near the context limit, tool search that loads tool definitions on demand, and programmatic tool calling for parallel and chained calls. It supports MCP, custom functions, and built-in web search.
- **Subagents built in:** A`multi_agent` setting delegates independent pieces of a task to parallel subagents with their own context. Early customer Ciridae reports its evaluation score rising from 0.71 to 0.85 and a 4x latency reduction.
- **Sandbox choice:** Agents run in OpenAI-hosted sandboxes, on your own infrastructure, or through partners Blaxel, Cloudflare, Daytona, DigitalOcean, E2B, Modal, Oracle, Runloop, and Vercel.
- **Open-source foundation:** The harness underneath is the open-source Codex harness, so developers can read the logic that coordinates calls, tools, and context.
- **Pricing and limits:** There are no extra fees beyond tokens and tools. Data residency is US-only, and Zero Data Retention is not supported yet. The HN thread passed 330 points.