Mode in Claude Code that fans out many sub-agents. The creator used it with max thinking effort for full code reviews, and it found many bugs and security issues. It is very expensive but now finishes in about 10 minutes instead of 60 to 120.
Ultra Code
by Anthropic
Claude Code mode that fans out many parallel sub-agents for deep tasks such as full codebase reviews.
Updated EU status checked Last shown
Ultra Code at a glance
- What is Ultra Code?
- Ultra Code is an agent from Anthropic. Claude Code mode that fans out many parallel sub-agents for deep tasks such as full codebase reviews.
- Who makes Ultra Code?
- Ultra Code is made by Anthropic.
- Does Ultra Code work in the EU?
- Unverified. No explicit evidence found. Last verified 14 Nov 2026.
- How much does Ultra Code cost?
- Pricing unknown. Described by a creator as very expensive; no official pricing found. Usage runs through Claude Code. Licence: Proprietary.
- Where was Ultra Code shown?
- In 2 creator videos by Matthew Berman. Most recently by Matthew Berman on 9 Nov 2026, at 21:10 in “Fable 5.2 solved all software...”.
Sources: creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.
Ultra Code is a mode in Anthropic's Claude Code, the agentic coding tool, that distributes work across many sub-agents running in parallel. No official documentation page could be retrieved, so the details below rest on a single creator account and should be treated as provisional.
According to a YouTube creator (Matthew Berman), it was used with maximum thinking effort to run full code reviews, where it surfaced many bugs and security issues. The same review reportedly finished in about 10 minutes, compared with 60 to 120 minutes previously, because the work is parallelised.
The creator described it as very expensive, since many sub-agents each consume tokens. It is accessed through Claude Code, so it is aimed at developers working on existing codebases. Exact pricing, plan requirements and limits are not confirmed here.
- Fans out work to many parallel sub-agents
- Works with maximum thinking effort
- Used for full codebase reviews, including bug and security findings
- Reported to cut review time to about 10 minutes from 60-120
- High token cost reported
As seen on YouTube
Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.
Workflows feature that starts a planning agent delegating to potentially hundreds of parallel sub-agents. Creator watched 100+ sub-agents run on a codebase review, burning a lot of tokens.
Similar AI tools
Closest neighbours by meaning, not just by category.
Claude Code
App · Anthropic
Anthropic's Claude assistant and agentic coding tool, with chat apps, Projects, skills, connectors and Claude Code.
Dynamic Workflows
App · Anthropic
Claude Code feature where Claude writes orchestration scripts that run tens to hundreds of parallel subagents with verification.
Claude Code Auto Mode
App · Anthropic
Claude Code permission mode where a classifier approves safe tool calls and blocks risky ones instead of prompting.
Claude Code Agent View
App · Anthropic
Claude Code CLI view that lists and manages multiple parallel Claude Code sessions, showing which need input or are done.
Claude Code Session Messaging
App · Anthropic
Claude Code feature that lets separate Claude Code sessions send each other messages and summaries to share context.
Claude Code in-app browser
App · Anthropic
In-app browser in the Claude Code desktop app with element annotation for requesting edits to web pages.