Claude Opus 5.5 in Claude Code 2.1.288
skills off
Zoom, inline player and more ways to view
The model's own page, run in a sandboxed frame only when you ask (a page that needs its own origin opens in a new tab on a separate demo site). Phones get a full-screen player capped at 1.5x pixel density (2x on larger screens) so it stays smooth; the original page runs uncapped. It starts in its auto-tour (T switches to play mode on a keyboard).
Wall time 17m 28sCost $3.77 reported
Review
No verdict yet.
The full review
OK. Finished on its own after 17.5 min.
Scores
AI judge
Not scored
No rubric for this prompt
Visitor votes and reviewer grades are not available yet.
- AI judge
- AI judge: the test runner (model claude-opus-5-5), scoring by visual review of the 1600 px render against the pre-registered rubric. No AI-judge score exists for this result: there is no rubric for this prompt, so no score is shown and none is made up.
- Visitors
- Public voting with a blind reveal is coming: you score the output first, then see the other scores.
- Reviewer
- The reviewer's own grade appears here once it is added.
Scores compare results for the same prompt only. AI-judge boards rank runs within one prompt, equal totals sharing a rank; scores from different prompts are never averaged.
Downloads
Files actually published for this run. Original output, site-made previews and post-run conversions are labelled separately.
The exact prompt
agentic harness (verbatim prompt + deliverable line), 268 characters. Every result for this prompt
Show me what goes on inside an AI data centre. --- Deliverable: build this in the current folder as a self-contained web page that opens in a browser with index.html as the entry point (no build step; CDN imports are fine). Work autonomously and do not ask questions.
How it was made
Step by step
The prompt, every agent turn and tool call (tool name and a one-line summary; inputs and outputs are not shown), then the final message. The full machine-format transcript is published as a scrubbed download.
Show all 53 steps
Loading the timeline…
Download the full transcript Claude Code stream JSON, 372 KB. Scrubbed: anything that looks like a key, token, e-mail address or phone number is replaced, and account details and the test machine's file paths are removed.
Cost breakdown
Headline cost $3.77, as reported by Claude Code (its own bill for the run). The list-price estimate from the token counts is $3.77.
| Tokens by type | Tokens | $ per 1M | Cost |
|---|---|---|---|
| Input (uncached) | 72 | $4.00 | $0.0003 |
| Cache write | 152,212 | $5.00 | $0.7611 |
| Cache read | 4,402,005 | $0.20 | $0.8804 |
| Output | 106,369 | $20.00 | $2.1274 |
| of which reasoning | 37,275 | billed as output |
- Total computed from tokens
- $3.7691
- Total reported by the harness
- $3.7691
- Difference
- matches
Claude Code 2.1.288 reported its own total; token counts x list prices shown beside it.
Prices: Project price table (vendor list prices), 2026-09-22.
Run notes
- Prompt from the open-source collection awesome-opus5-5-videos (github.com/yihui-dev/awesome-opus5-5-videos, MIT License, (c) 2026 yihui-dev), originally posted by @mdaman010: https://x.com/mdaman010/status/2103845689713111079
- Prompt sent verbatim, plus one fixed deliverable line (build index.html in the run folder, no build step, CDN imports allowed, work autonomously). One scored attempt per harness; this is the 2nd run of this cell after an infra timeout (see next note).
- Run with no time or turn limit. An earlier attempt of this cell was stopped by our own 30-minute timeout on an overloaded machine while the agent was testing its page; that attempt is not shown (it was our infrastructure, not the model).
- Capture: 1280x720 headless Chromium, software WebGL (no GPU); capture fps 5 is a software-render figure, not a real-device frame rate. Console errors during capture: 0.
- Mode: headless (claude -p, stream-json)