Kimi K3 in Kimi Code CLI 2.1.0
skills off
Zoom, inline player and more ways to view
Wall time 27m 44sCost $1.68 estimated8.0/10 · Draft score — stored total; criterion breakdown unavailable
Review
Solid MacBook Pro in three-quarter view on a dark backdrop: notch, a vivid wallpaper with a window and dock, blue light spill, speaker grilles, a big trackpad, side ports and a front notch.
The full review
Solid MacBook Pro in three-quarter view on a dark backdrop: notch, a vivid wallpaper with a window and dock, blue light spill, speaker grilles, a big trackpad, side ports and a front notch. The keyboard is a flat unlabelled grid, the trackpad sits off-centre toward the right, and the aluminium shading is flat, with little edge highlight.
Verdict by the AI judge (provisional).
OK. Finished on its own after 27.7 min and produced macbook.svg.
Bugs
- keyboard is a flat grid
- trackpad off-centre
- flat aluminium
Scores
AI judge
8.0/10
Draft score — stored total; criterion breakdown unavailable
Visitor votes and reviewer grades are not available yet.
Draft score — stored total; criterion breakdown unavailable. the per-criterion marks could not be read, so the stored score is shown; the displayed total has not been verified through a criterion breakdown.
- 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. A draft score until the reviewer checks it. Rubric: Wave-1 SVG scoring rubric (AI Model Reviewer).
- 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 (save the file; self-render allowed), 373 characters. Every result for this prompt
Generate an SVG that looks like a realistic 3D render of a MacBook Pro 16" (opened, slight three-quarter view, screen glow, aluminum body). Use gradients and perspective to fake the 3D. Save it as macbook.svg in the current directory. You may render and inspect your own output (headless chromium is available at /usr/bin/chromium) and iterate on your own before finishing.
How it was made
Versions saved during the run
5 intermediate outputs, in the order the model wrote them.
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 transcript is 2.6 MB; it is published as a scrubbed, compressed download rather than shown on the page.
Show all 2 steps
Loading the timeline…
Download the full transcript transcript, 151 KB, gzip. 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 $1.68, estimated from the token counts at list price.
| Tokens by type | Tokens | $ per 1M | Cost |
|---|---|---|---|
| Input (uncached) | 82,919 | $3.00 | $0.2488 |
| Cache write | 0 | $0.00 | $0 |
| Cache read | 2,338,634 | $0.30 | $0.7016 |
| Output | 48,318 | $15.00 | $0.7248 |
| of which reasoning | not recorded | billed as output |
- Total computed from tokens
- $1.6751
- Total reported by the harness
- none
Tokens summed from every step.end usage in Kimi Code's wire.jsonl (main agent + any subagent wires): inputOther = uncached input, inputCacheRead, inputCacheCreation, output (reasoning is included in output, not split). Kimi Code prints no dollar figure: cost = tokens x Moonshot list price.
Prices: models.dev catalog as loaded by OpenCode 1.18.32 (read 2026-09-23) (Moonshot kimi-k3 list price), 2026-09-23.
Run notes
- Playcode MacBook SVG Benchmark prompt verbatim, except its last sentence ("Respond with ONLY the SVG markup, no explanation.") is swapped for the save-to-file harness suffix, as the prompt research recommends for harness runs.
- Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
- Isolation: bwrap: private /tmp, /run, PID/IPC/UTS ns; HOME, TMPDIR and the harness home (sessions, config copy) inside this run dir; only /usr, /etc, the harness binary dir(s) (ro) and this run dir bound; the project files, results/, other runs, /tmp/hx_* not mounted