Codex Subagents
Parallel custom agents for complex tasks
Codex Subagents launched on March 17, 2026 with 300 votes, #84 of 23,380 launches that month and more than 99% of that year's launches.
In plain words
Codex Subagents enables users to create specialized AI workers that operate in parallel on complex coding tasks. The tool uses custom TOML-configured agents with defined roles, such as explorers and reviewers, to execute multi-step workflows while maintaining context integrity. Designed for developers handling intricate projects, this productivity tool allows teams to distribute work across isolated agents, reducing the context degradation that typically occurs in sequential processing.
written from the facts on this page · September 2026
From the sources
Codex now supports subagents, allowing you to spawn specialized, parallel AI workers for complex coding tasks. By defining custom TOML agents with isolated roles (like explorers and reviewers), you can execute multi-step workflows without context rot.
Use subagents in ChatGPT and Codex, and configure custom Codex agents
ChatGPT Work and Codex can run subagent workflows by spawning specialized agents in parallel and then collecting their results in one response. This can be particularly helpful for complex tasks that are highly parallel, such as codebase exploration or implementing a multi-step feature plan. In local Codex clients, you can also define custom agents with different model configurations and instructions for different tasks. ChatGPT Work exposes subagent workflows and activity to eligible accounts. Current Codex releases enable subagent workflows by default. Subagent activity appears in the ChatGPT desktop app, Codex CLI, and the IDE extension. Because each subagent does its own model and tool…from learn.chatgpt.com
Does a similar job
all alternatives →
Parallel CodeMar 2026 · parallelcode.app · ▲159I kept running into a structural problem with AI coding tools. You try one agent. It makes changes. You try another. Now you're manually diffing branches and cleaning up experimental commits. The tools are powerful, but the workflow isn’t designed for comparison. So I built Parallel Code. It runs Claude Code, Codex, and Gemini CLI natively (no API wrapper, no feature abstraction), but isolates each agent in: its own terminal its own Git worktree its own feature branch The workflow becomes: Spawn N worktrees Assign one AI agent per worktree Let them implement independently Compare diffs Merge…
20+ Claude Code agents coordinating on real work (open source)Feb 2026 · github.com · ▲53Single-agent LLMs suck at long-running complex tasks. We’ve open-sourced a multi-agent orchestrator that we’ve been using to handle long-running LLM tasks. We found that single LLM agents tend to stall, loop, or generate non-compiling code, so we built a harness for agents to coordinate over shared context while work is in progress. How it works: 1. Orchestrator agent that manages task decomposition 2. Sub-agents for parallel work 3. Subscriptions to task state and progress 4. Real-time sharing of intermediate discoveries between agents We tested this on a Putnam-level math problem, but the…
More ai this month
the category →
Paintings by AI models: programs run through a simulation of oil paint.
AI · 20d ago · stillwet.art
PixVerse R2▲381A real-time world model you can explore and change
AI · 28d ago · world.pixverse.video
Source: Product Hunt launch ↗
Launched alongside, March 2026
the whole month →
Switch from ChatGPT to Claude with import memory feature
AI · Mar 2026 · claude.com
Littlebird▲709Littlebird remembers everything across your meetings, messages, docs, and apps, then takes action so you can recall any detail and never lose context.
AI · Mar 2026 · littlebird.ai
