ultracode and Effort Controls
As of 2026-10-08 (Claude Code 2.1.284 and later), ultracode is its own toggle in /effort, not an effort level: with it on, Claude orchestrates dynamic workflows for substantive tasks at whichever effort level the session runs at. max is the deepest reasoning level. Below: the levels and how ultracode differs, per the official docs.
Point this tool at the QCode endpoint — one key for Claude, GPT and Chinese models.
- Billed per token — live rates on /models
- Pay by card (Visa / Mastercard / AMEX), Apple Pay, Google Pay or crypto
- Self-service: top up and activate instantly
The effort levels
In Claude Code, Opus 5.5, Sonnet 5.5, Opus 5, Sonnet 5, Opus 4.8 and Opus 4.7 all support low, medium, high, xhigh and max; per the docs, each level trades token spend against capability.
| Level | In Claude Code | Best for |
|---|---|---|
| High | /effort high |
Work where verification matters or edge cases are likely, such as fixing a bug in an existing codebase. The default on every model except Opus 5.5, Sonnet 5.5 and Opus 4.7. |
| Extra | /effort xhigh |
Deeper reasoning at higher token spend; the default on Opus 4.7. Ultracode is only available on models that support xhigh. |
| Max | /effort max |
The deepest reasoning level, for hard problems you want Claude to work through without you, such as finding security vulnerabilities. The docs warn max may show diminishing returns and is prone to overthinking, so test before adopting it broadly; unless set via the CLAUDE_CODE_EFFORT_LEVEL environment variable, Claude Code applies max to the current session only. |
As of 2026-10-08: in Claude Code, Opus 5.5 and Sonnet 5.5 default to medium (day-to-day engineering work with a clear scope, such as implementing a new feature), Opus 4.7 defaults to xhigh, and every other model that supports effort defaults to high.
What is ultracode?
A Claude Code setting rather than a model effort level; since 2.1.284 it stays on at any effort level.
Per the official docs: with ultracode on, Claude orchestrates dynamic workflows for substantive tasks at whichever effort level the session runs at. Turning it on or off with /effort or the ultracode setting leaves the effort level unchanged; the --effort ultracode flag and the Agent SDK effortLevel: "ultracode" value turn it on and also set the level to xhigh. Before 2.1.284, turning on ultracode set the session to xhigh and picking another level turned it off. Now, picking a level in the /effort slider or the /model picker leaves ultracode as it was; keeping ultracode on at levels other than xhigh requires Claude Code 2.1.284 or later.
- •Run /effort ultracode to turn it on for the current session and /effort ultracode off to turn it off; in the /effort slider, press Tab to flip the Ultracode toggle.
- •Launch with claude --effort ultracode to start the session at xhigh with ultracode on, or set "ultracode": true in a settings file.
- •It is unavailable when workflows are turned off or the model doesn't support xhigh.
/effort ultracode · /effort ultracode off · claude --effort ultracodeWhich level should you use?
Match effort to the task — higher isn't always better, because it costs more time and tokens.
Routine work → medium / high
Opus 5.5 and Sonnet 5.5 default to medium, which fits day-to-day engineering with a clear scope; use high where verification matters or edge cases are likely, such as fixing a bug in existing code.
Hard problems → xhigh / max
Tricky debugging, architecture, or anything where a wrong answer is expensive. xhigh reasons deeper at higher token spend; max suits hard problems Claude works through without you, but the docs warn of diminishing returns, so test first.
Long async runs → ultracode
Big, multi-stage tasks you let run in the background. ultracode plans a dynamic workflow for each substantive task and, since 2.1.284, stays on at any effort level instead of forcing xhigh.
ultracode and dynamic workflows
To run a single task as a workflow without changing the session's effort level, include the keyword ultracode in your prompt; with /effort ultracode, Claude plans a workflow for every substantive task in the session.
Read the dynamic workflows guide →Fast Mode pricing
Per the official price list as of 2026-10-08: Opus 5.5 Fast mode is $8 per million input and $40 per million output tokens, 2x its standard $4 / $20 (converted from the official price list); Opus 5 and Opus 4.8 Fast mode is $10 / $50. Fast mode is still a research preview; the price list says it is available on the Claude API (first-party) only, not on Claude Platform on AWS or partner-operated cloud platforms.
Higher effort costs more — run it on the right access
xhigh, max, and ultracode reason longer and, with workflows, spawn many agents — so they use more tokens and draw down rate limits faster than routine work.
Sustained high-effort sessions with QCode
As of 2026-10-08, claude-opus-5-5, claude-sonnet-5-5 and claude-opus-4-8 are all callable with one QCode key. Billing is per token, with per-model prices on /models; xhigh, max and ultracode use more tokens, so estimate usage before long runs.
Run ultracode on Opus 5.5 in Claude Code
Sign up for QCode and call claude-opus-5-5 and claude-sonnet-5-5 with one key. Billing is per token; per-model prices are on /models.