AI Model Reviewer

DeepSeek V4.1 Flash in OpenCode 1.18.32

skills off

← Open in the gallerySame prompt, all models

DeepSeek V4.1 Flash's output for BMW M4 Competition, side view (flat vector)

▶ Open live demo

Zoom, inline player and more ways to view
DeepSeek V4.1 Flash's SVG for BMW M4 Competition, side view (flat vector)

OKBMW M4 Competition, side view (flat vector)max effortNeutral harness

Wall time 39m 51sCost $0.24 reported

Open the SVG filePoster render (WebP)

Review

No verdict yet.

The full review

OK. Finished on its own after 39.9 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 (save the file; self-render allowed), 441 characters. Every result for this prompt

Create a clean SVG illustration of a BMW M4 Competition in a side view. The image must follow a 4:3 aspect ratio. Use the original factory color of the BMW M4 Competition. Focus on coupe proportions and sporty stance. Use vector shapes without gradients. Save it as bmw_m4.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

Wall time 39m 51sFirst artifact at 3m 32s32 saved versions

Versions saved during the run

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

Version 32 of 32
  1. Version 1 at 3m 42s thumbnailv1 3:42
  2. Version 2 at 5m 23s thumbnailv2 5:23
  3. Version 3 at 7m 42s thumbnailv3 7:42
  4. Version 4 at 11m 42s thumbnailv4 11:42
  5. Version 5 at 13m 42s thumbnailv5 13:42
  6. Version 6 at 16m 31s thumbnailv6 16:31
  7. Version 7 at 19m 34s thumbnailv7 19:34
  8. Version 8 at 21m 39s thumbnailv8 21:39
  9. Version 9 at 22m 28s thumbnailv9 22:28
  10. Version 10 at 22m 38s thumbnailv10 22:38
  11. Version 11 at 23m 27s thumbnailv11 23:27
  12. Version 12 at 24m 44s thumbnailv12 24:44
  13. Version 13 at 25m 27s thumbnailv13 25:27
  14. Version 14 at 27m 00s thumbnailv14 27:00
  15. Version 15 at 27m 38s thumbnailv15 27:38
  16. Version 16 at 28m 28s thumbnailv16 28:28
  17. Version 17 at 28m 56s thumbnailv17 28:56
  18. Version 18 at 29m 50s thumbnailv18 29:50
  19. Version 19 at 30m 35s thumbnailv19 30:35
  20. Version 20 at 31m 41s thumbnailv20 31:41
  21. Version 21 at 32m 14s thumbnailv21 32:14
  22. Version 22 at 32m 42s thumbnailv22 32:42
  23. Version 23 at 33m 04s thumbnailv23 33:04
  24. Version 24 at 33m 53s thumbnailv24 33:53
  25. Version 25 at 34m 16s thumbnailv25 34:16
  26. Version 26 at 35m 54s thumbnailv26 35:54
  27. Version 27 at 36m 55s thumbnailv27 36:55
  28. Version 28 at 37m 18s thumbnailv28 37:18
  29. Version 29 at 37m 53s thumbnailv29 37:53
  30. Version 30 at 38m 20s thumbnailv30 38:20
  31. Version 31 at 38m 42s thumbnailv31 38:42
  32. Version 32 at 39m 10s thumbnailv32 39:10

No transcript was delivered with this result.

Cost breakdown

Headline cost $0.24, as reported by OpenCode (its own bill for the run). The list-price estimate from the token counts is $0.24.

Tokens by typeTokens$ per 1MCost
Input (uncached)89,854$0.15$0.0135
Cache write0$0.15$0
Cache read31,525,120$0.00$0.0946
Output224,149$0.60$0.1345
of which reasoning151,965billed as output
Total computed from tokens
$0.2425
Total reported by the harness
$0.2425
Difference
matches

OpenCode 1.18.32 reported its own total; token counts x list prices shown beside it. The harness reports reasoning tokens separately from output; here output includes them (72,184 visible + 151,965 reasoning), and both bill at the output rate.

Prices: https://api-docs.deepseek.com/quick_start/pricing, 2026-09-24.

Run notes

one prompt, agentic harness (self-review allowed)labelled re-run (of bmw_m4_side__opencode__deepseek-flash)

  • frozen prompt (from_run): recovery/prompts_recovered/bmw_m4_side.from_run.txt: byte-exact first user message of a real run
  • plain OpenCode config copy (providers only; no ECC instructions/agents)
  • opencode export written to a file (not piped)
  • Run on a GitHub Codespace (codespaces lane of box-fleet); model calls reached the providers through a credential relay on the home machine (no keys in the codespace).
  • Transcript redaction: base64 thinking signatures (a known secret-scanner false positive) are replaced with "[signature omitted]" by the engine offload before push; nothing else in the model output is altered.
  • Mode: interactive TUI in tmux (160x48), inside bubblewrap sandbox
  • Isolation audit: outside paths seen