AI Model Reviewer

Kimi K3 in Kimi Code CLI 2.1.0

skills off

← Open in the gallerySame prompt, all models

Poster frame of Kimi K3's page
Open the original page

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).

Recorded from the live demo (stepped clock). Preview video recorded by this site from the live demo. The moving picture above is that preview; the button runs the real page.

Stopped by usGPU fluid simmax effortNeutral harness

Wall time 57m 02sCost $2.29 estimated

Snapshot 8 of 8 saved during Kimi K3's run
  1. Snapshot 1 of 8 saved during Kimi K3's run
  2. Snapshot 2 of 8 saved during Kimi K3's run
  3. Snapshot 3 of 8 saved during Kimi K3's run
  4. Snapshot 4 of 8 saved during Kimi K3's run
  5. Snapshot 5 of 8 saved during Kimi K3's run
  6. Snapshot 6 of 8 saved during Kimi K3's run
  7. Snapshot 7 of 8 saved during Kimi K3's run
  8. Snapshot 8 of 8 saved during Kimi K3's run
Still snapshots saved during the run (the moving picture above was recorded from the live demo): 8 distinct frames.

Generation timelapse

Poster for generation timelapse

Generation timelapse (the agent's screen during the run, sped up). This is not the finished video. Download video (1.2 MB).

Review

No verdict yet.

The full review

Stopped by us. Stopped by us (box row cap reached (3704s); stopped by the box lane so the run ships as partial); not a model failure.

Attempt 2 of 2

Every try of this model in this harness on this prompt, oldest first. This run is the cell's result.

  1. Attempt 1 ended in an infrastructure failure, not a model result2026-09-26
  2. Attempt 2 Stopped by us this run57m 02s$2.292026-09-26

Scores

Not scored: only runs that finished on their own are scored.

Downloads

Files actually published for this run. Original output, site-made previews and post-run conversions are labelled separately.

The exact prompt

as run, 664 characters. Every result for this prompt

Build a GPU-accelerated fluid simulation in a single HTML file using WebGL, no external libraries. Dragging the mouse or a finger should push colorful dye through the fluid, with a soft glow. Make it look stunning and run smoothly.

Build it as a self-contained index.html in the current directory, with WebGL/WebGL2 shaders you write yourself; no libraries and no CDNs; interactive with mouse or touch input. You may serve it locally and open it in headless chromium (/usr/bin/chromium, software WebGL, no GPU) 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.

How it was made

Wall time 57m 02sFirst artifact at 12m 00s19 saved versions

Versions saved during the run

19 intermediate outputs, in the order the model wrote them.

Version 19 of 19
  1. Version 1 at 12m 04s thumbnailv1 12:04
  2. Version 2 at 19m 53s thumbnailv2 19:53
  3. Version 3 at 33m 21s thumbnailv3 33:21
  4. Version 4 at 33m 31s thumbnailv4 33:31
  5. Version 5 at 35m 17s thumbnailv5 35:17
  6. Version 6 at 35m 52s thumbnailv6 35:52
  7. Version 7 at 37m 13s thumbnailv7 37:13
  8. Version 8 at 37m 39s thumbnailv8 37:39
  9. Version 9 at 37m 54s thumbnailv9 37:54
  10. Version 10 at 38m 09s thumbnailv10 38:09
  11. Version 11 at 38m 24s thumbnailv11 38:24
  12. Version 12 at 38m 34s thumbnailv12 38:34
  13. Version 13 at 39m 50s thumbnailv13 39:50
  14. Version 14 at 41m 25s thumbnailv14 41:25
  15. Version 15 at 45m 43s thumbnailv15 45:43
  16. Version 16 at 47m 54s thumbnailv16 47:54
  17. Version 17 at 48m 04s thumbnailv17 48:04
  18. Version 18 at 55m 55s thumbnailv18 55:55
  19. Version 19 at 56m 20s thumbnailv19 56:20

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 2.9 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, 173 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 $2.29, estimated from the token counts at list price.

Tokens by typeTokens$ per 1MCost
Input (uncached)94,116$3.00$0.2823
Cache write0$3.00$0
Cache read4,109,162$0.30$1.2327
Output51,918$15.00$0.7788
of which reasoningnot recordedbilled as output
Total computed from tokens
$2.2939
Total reported by the harness
none

Computed from harness-reported token counts x list prices. Kimi Code prints no dollar figure.

Prices: https://platform.kimi.ai/docs/pricing/chat, 2026-09-24.

Run notes

one prompt, agentic harness (self-review allowed)labelled re-run (of gpu_fluid_sim__kimi-code__kimi-k3)

  • frozen prompt (written): engine/prompts_written/gpu_fluid_sim.txt: Voxyz's own prompt verbatim (x.com/Voxyz_ai/status/2103088468762960378) + our fixed inputs (self-contained index.html, own WebGL/WebGL2 shaders, no libraries/CDNs, mouse/touch input) (supervisor 725c3fc2 05:53Z; zero results)
  • Kimi Code on Node 22.23.3 through the per-run credential relay
  • Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
  • Isolation audit: outside paths seen