TASK 08 / HTML · V1 · Game

Single-File Breakout Game

单文件打砖块游戏

4 runs · 08.A.01 · HTML

Compare works →
TASK 08 · 08.A.01 · HTMLSingle-File Breakout Game
Open independent copy ↗
Single-File Breakout Game

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

Create a polished, playable Breakout game in a single HTML file.
TRANSLATION

Chinese translation

在单个 HTML 文件中创建一个精致且可玩的打砖块游戏。

Expected outcome

The deliverable should be one self-contained HTML file containing all markup, styles, scripts, and required assets, runnable without a build step. The game should include a paddle, ball, bricks, collision handling, scoring, lives or a loss condition, and coherent start, pause, restart, win, and fail states. Controls should feel responsive, physics should be consistent, and fast motion should not routinely tunnel through bricks or the paddle. “Polished” should also mean a clear HUD, cohesive art direction, restrained animation and sound, explicit state messaging, responsive layout, and basic accessible input support.

Additional verification guidance

  • Open the single file offline and confirm there are no missing dependencies, console exceptions, or server-only requirements.
  • Exercise start, launch, movement, brick hits, scoring, ball loss, pause, resume, game over, restart, and victory flows end to end.
  • Test wall, paddle, and brick collisions across frame rates and ball speeds, checking for stalls, endless horizontal paths, or obvious tunneling.
  • Test keyboard, mouse, and touch where supported, and verify a complete playfield and readable HUD on narrow and wide screens.
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-08-r1Muse Spark 1.3run-muse-spark-1-3-task-08-r1GPT-5.6 Solrun-gpt-5-6-sol-task-08-r1K3run-k3-task-08-r1
Runtime11:571:187:032:53
API calls466135
Tool calls54564
Failures / retries2000
Token throughput3M94K546.2K160.7K
AI review92/10088/10083/10092/100
RUN RECORDS

Run records

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

08.A.01 · run-deepseek-v4-1-flash-exp-0910-task-08-r1DeepSeek-V4.1-Flash● Archived
Runtime
11:57
Token throughput
3M
API calls
46
View archive record →
08.B.01 · run-muse-spark-1-3-task-08-r1Muse Spark 1.3● Archived
Runtime
1:18
Token throughput
94K
API calls
6
View archive record →
08.C.01 · run-gpt-5-6-sol-task-08-r1GPT-5.6 Sol● Archived
Runtime
7:03
Token throughput
546.2K
API calls
13
View archive record →
RELATED TASKS

Related tasks

View all tasks →