diri
Measured by ADEMark
ADEMark ran the same tests on 15 ADEs. Results for diri 0.11.2, measured up to 12 November 2026 on a Mac mini M4, 16 GB, macOS 27 · Claude Opus 5.5, effort high. Data CC BY 4.0, shown unchanged.
Speed
- Cold start
- 2.26 s6th of 12median of 6 runs, 2.21 s to 2.29 s · Bare CLI 0.69 s
- New thread, ready to type
- 663 ms10th of 10median of 10 runs, 620 ms to 3,440 ms · Bare CLI 375 ms
- First word on screen
- 256 ms1st of 11median of 9 runs, 241 ms to 276 ms · Bare CLI 220 ms
- Typing lag, 5 agents streaming
- 72 ms9th of 111 run · Bare CLI 79 ms
- Open a 2,000-message thread
- 88 ms2nd of 11median of 5 runs, 68 ms to 106 ms · Bare CLI 203 ms
Weight
- Memory at rest, 5 threads
- Not measuredADEMark: it starts its agents detached, so we cannot yet tell them from other Claude processes
- Idle CPU
- 0.55%2nd of 131 run · Bare CLI 0.37%
- Memory under load
- Not measuredADEMark: it starts its agents detached, so we cannot yet tell them from other Claude processes
Tokens
- Tokens to answer “hi”
- 39.5K tokens6th of 13median of 3 runs, 39.3K tokens to 39.5K tokens · Bare CLI 38.2K tokens
- Tokens for one real question
- 175.8K tokens10th of 131 run · Bare CLI 168.3K tokens
Build
- Team build: hidden tests passed
- Not measured
Place counts only apps with a value; lower is better except passed tests. ADEMark publishes no overall score and treats close medians as ties, so read places as an order, not a verdict. The Bare CLI is Claude Code in a terminal, run as an unranked control. Method and every test on ademark.dev · Imported data · How we use it