TASK 11 / HTML · V1 · Productivity tool

2D Floor-Plan Editor

2D 户型图编辑器

4 runs · 11.A.01 · HTML

Compare works →
TASK 11 · 11.A.01 · HTML2D Floor-Plan Editor
Open independent copy ↗
2D Floor-Plan Editor

Preview note: Embedded storage and downloads can be restricted by browser sandboxing; use the standalone view for full behavior.

ORIGINAL PROMPT

Original prompt

PROMPT/Phase1_TEST_PROMPTS_BILINGUAL.md · The original prompt is kept verbatim

Source ↗
ORIGINAL PROMPT

Original prompt

Build a simple 2D floor-plan editor where I can draw walls, add doors and windows, drag objects, and see room dimensions.
TRANSLATION

Chinese translation

构建一个简单的 2D 户型图编辑器,让我可以绘制墙体、添加门窗、拖动物体并查看房间尺寸。

Expected outcome

The deliverable should be an interactive 2D floor-plan editor with at least four core capabilities: wall drawing, doors and windows attached to walls, object placement and dragging, and room-dimension display. Coordinates, scale, and snapping should be consistent. Connected wall endpoints should form recognizable enclosed rooms, and openings should remain attached to their wall when moved. Object dragging should preserve selection and positional accuracy. Dimensions should state their units and measurement convention and update live when geometry changes. The UI should make tools and selection states discoverable and provide recovery from mistakes.

Additional verification guidance

  • Draw a rectangle and a multi-room plan with T-junctions; verify endpoint snapping, closed-loop detection, room count, width, height, and area.
  • Add doors and windows to different walls and drag them; ensure they do not detach, cross endpoints, or become invalid after wall resizing.
  • Add, select, drag, and delete objects; verify hit testing, layering, coordinates, snapping, and full state restoration through undo/redo.
  • Repeat edits after zooming and panning to validate screen-to-world conversion; check units, live dimension updates, save/load, and malformed-data handling.
THREE INTERPRETATIONS

Same task, three model interpretations

All three results use the same width and the same preview ratio; no model is marked as best.

COMPARISON METRICS

Comparison metrics

Cross-model comparison must note differences in harness, budget and isolation.

MetricDeepSeek-V4.1-Flashrun-deepseek-v4-1-flash-exp-0910-task-11-r1Muse Spark 1.3run-muse-spark-1-3-task-11-r1GPT-5.6 Solrun-gpt-5-6-sol-task-11-r1K3run-k3-task-11-r1
Runtime12:101:439:0418:54
API calls81122219
Tool calls8012718
Failures / retries4002
Token throughput7.8M264.2K1M1.2M
AI review97/10084/10064/10087/100
RUN RECORDS

Run records

Two runs of the same task version can be compared side by side.

11.A.01 · run-deepseek-v4-1-flash-exp-0910-task-11-r1DeepSeek-V4.1-Flash● Archived
Runtime
12:10
Token throughput
7.8M
API calls
81
View archive record →
11.B.01 · run-muse-spark-1-3-task-11-r1Muse Spark 1.3● Archived
Runtime
1:43
Token throughput
264.2K
API calls
12
View archive record →
11.C.01 · run-gpt-5-6-sol-task-11-r1GPT-5.6 Sol● Archived
Runtime
9:04
Token throughput
1M
API calls
22
View archive record →
RELATED TASKS

Related tasks

View all tasks →