The wiki is available in English and Simplified Chinese. Showing English.
Codex
Yes. Subagent workflows consume more tokens than single-agent runs.
Each subagent does its own model and tool work. Cloud tasks and local sessions share one allowance, and a cloud chat can use more of it than a local message. Delegated workers are billed like any other Codex activity. OpenAI does not document a separate cap on concurrent tasks.
What the official docs say
Because each subagent does its own model and tool work, subagent workflows consume more tokens than comparable single-agent runs.Cloud chats on ChatGPT plans use GPT-5.6 Sol and may use more of your allowance than local messages.These rates apply to supported ChatGPT Work and Codex activity, including local tasks, cloud tasks, automations, code review, auto review, and delegated workers.Claude
Yes. Subagents, agent teams and parallel web sessions all draw from the same limits.
/usage attributes recent usage to subagents, skills, plugins and MCP servers. Claude Code on the web shares limits with everything else in your account, and parallel tasks consume proportionately more. Agent teams use roughly 7x the tokens of a standard session because each teammate is its own Claude instance. A large fanout can exhaust the weekly allowance before the session window resets.
What the official docs say
On a Pro, Max, Team, or Enterprise plan, /usage also shows a breakdown of what counts against your plan limits: Attribution: recent usage attributed to skills, subagents, plugins, and individual MCP servers, each shown as a percentage of the total.Rate limits: Claude Code on the web shares rate limits with all other Claude and Claude Code usage within your account. Running multiple tasks in parallel consumes more rate limits proportionately. There is no separate compute charge for the cloud VM.Agent teams use approximately 7x more tokens than standard sessions when teammates run in plan mode, because each teammate maintains its own context window and runs as a separate Claude instance.A single burst of heavy activity, such as a large workflow fanout, can exhaust the weekly allowance before the session window resets.In practice
- Route subagents to a smaller model when the task allows; the main window pays for every subagent turn.
- Parallel sessions do not get separate limits anywhere; they only spend the same one faster.
Quotes checked against the live docs on 2026-09-15.