Claude Sonnet 5.5 in Claude Code 2.1.293
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 8m 50sCost $0.84 reported
Review
No verdict yet.
The full review
OK. Finished on its own after 8.8 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
as run, 1,058 characters. Every result for this prompt
Build a single self-contained index.html (Three.js from a CDN) showing interactive soap bubbles. Each bubble uses a custom shader for thin-film iridescence (colour changes with view angle and film thickness), plus a fresnel rim and a reflection from a procedural environment. Click to blow a new bubble from the cursor. Bubbles drift on a gentle noise wind field, bounce softly off each other, and pop with a small particle burst when they touch the floor or are clicked. Include a UI slider for film thickness, wind strength and bubble count (max 60), and an FPS readout. One file, no build step, no external assets. Save the main page as index.html in the current directory (plus any local asset files you create), with no build step. You may serve it locally and open it in headless chromium (/usr/bin/chromium, software WebGL, no GPU, no WebGPU) to screenshot and inspect your own output, and iterate on your own before finishing. Close any browser you open as soon as you've taken your screenshot. Finish with a short README.md describing the controls.
How it was made
Versions saved during the run
5 intermediate outputs, in the order the model wrote them.
No transcript was delivered with this result.
Cost breakdown
Headline cost $0.84, as reported by Claude Code (its own bill for the run). The list-price estimate from the token counts is $0.72.
| Tokens by type | Tokens | $ per 1M | Cost |
|---|---|---|---|
| Input (uncached) | 44 | $2.00 | $0.0001 |
| Cache write | 72,115 | $2.50 | $0.1803 |
| Cache read | 1,221,416 | $0.10 | $0.1221 |
| Output | 41,581 | $10.00 | $0.4158 |
| of which reasoning | 17,373 | billed as output |
- Total computed from tokens
- $0.7183
- Total reported by the harness
- $0.8405
- Difference
- Harness-reported $0.8405 vs list-price estimate $0.7183 (-14.5%); the harness figure is the headline.
Claude Code 2.1.293 reported its own total; token counts x list prices shown beside it.
Prices: https://www.anthropic.com/claude/sonnet (input/output) and https://platform.claude.com/docs/en/about-claude/pricing (cache read 5% of input; 5-minute cache write 1.25x input), 2026-10-08.
Run notes
- frozen prompt (frozen): exact prompt_text of None on the runs branch (sha256 6d8642432e92); frozen prompt verbatim (prompts_queue, branch alpha/prompts-queue-20261008 @ 0b2dad4e, sha256 ab8d74fb20620089a953190c6143de3cc065f292811a26c848bb396b81836605), followed by the standard agent-harness
- no TUI: the harness ran in its non-interactive mode, so there are no TUI screenshots; the timelapse is built from version renders only
- fresh default harness config (no settings, agents or instruction files copied)
- infra retries: 0
- in the published transcript the sandbox HOME directory is shown as <HOME>
- Mode: headless print mode (claude -p, stream-json), inside bubblewrap sandbox
- Isolation audit: none
- Renderer: Three.js r160 (three@0.160.0 ES module from cdn.jsdelivr.net, imported at index.html line 30) drawing with WebGLRenderer (line 87). Needs WebGL; no WebGPU needed.
- Card video: none (the run's video.mp4 is not H.264 yuv420p with 30+ frames under 24 MB). Poster: the run's own render.png.