concepts/subagents.md Codex Docs, 2026-03-17 00:33 UTC → 2026-05-12 01:59 UTC

1# Subagents1# Subagents

2 2 

3Codex can run subagent workflows by spawning specialized agents in parallel so3Codex can run subagent workflows by spawning specialized agents in parallel so

4they can explore, execute, or analyze work concurrently.4they can explore, tackle, or analyze work concurrently.

5 5 

6This page explains the core concepts and tradeoffs. For setup, agent configuration, and examples, see [Subagents](https://developers.openai.com/codex/subagents).6This page explains the core concepts and tradeoffs. For setup, agent configuration, and examples, see [Subagents](https://developers.openai.com/codex/subagents).

7 7 

65 65 

66If you don't pin a model or `model_reasoning_effort`, Codex can choose a setup66If you don't pin a model or `model_reasoning_effort`, Codex can choose a setup

67that balances intelligence, speed, and price for the task. It may favor67that balances intelligence, speed, and price for the task. It may favor

68`gpt-5.3-codex-spark` for fast scans or a higher-effort `gpt-5.4`68`gpt-5.4-mini` for fast scans or a higher-effort `gpt-5.5` configuration for

69configuration for harder reasoning. When you want finer control, steer that69more demanding reasoning when that model is available. When you want finer

70choice in your prompt or set `model` and `model_reasoning_effort` directly in70control, steer that choice in your prompt or set `model` and

71the agent file.71`model_reasoning_effort` directly in the agent file.

72 72 

73For most tasks in Codex, start with `gpt-5.4`. Use `gpt-5.3-codex-spark` when73For most tasks in Codex, start with `gpt-5.5` when it is available. Continue

74you want a faster option for lighter subagent work.74 using `gpt-5.4` during the rollout if `gpt-5.5` is not yet available. Use

75 `gpt-5.4-mini` when you want a faster, lower-cost option for lighter subagent

76 work. If you have ChatGPT Pro and want near-instant text-only iteration,

77 `gpt-5.3-codex-spark` remains available in research preview.

75 78 

76### Model choice79### Model choice

77 80 

78- **`gpt-5.4`**: Start here for most agents. It combines strong coding, reasoning, tool use, and broader workflows. The main agent and agents that coordinate ambiguous or multi-step work fit here.81- **`gpt-5.5`**: Start here for demanding agents when it is available. It is strongest for ambiguous, multi-step work that needs planning, tool use, validation, and follow-through across a larger context.

79- **`gpt-5.3-codex-spark`**: Use for agents that favor speed over depth, such as exploration, read-heavy scans, or quick summarization tasks. It works well for parallel workers that return distilled results to the main agent.82- **`gpt-5.4`**: Use this when `gpt-5.5` is not yet available or when a workflow is pinned to GPT-5.4. It combines strong coding, reasoning, tool use, and broader workflows.

83- **`gpt-5.4-mini`**: Use for agents that favor speed and efficiency over depth, such as exploration, read-heavy scans, large-file review, or processing supporting documents. It works well for parallel workers that return distilled results to the main agent.

84- **`gpt-5.3-codex-spark`**: If you have ChatGPT Pro, use this research preview model for near-instant, text-only iteration when latency matters more than broader capability.

80 85 

81### Reasoning effort (`model_reasoning_effort`)86### Reasoning effort (`model_reasoning_effort`)

82 87