Nicheloom

Market intelligence for builders — see what's gaining traction before it's crowded.

Pac-Bench

How well can models one-shot a Pac-Man game?

Details

External ID
49885493
Source
HN
Company
—
Product
Pac-Bench
Website domain
github.io
Launched
Sept. 28, 2026
Cohort
—
Upvotes
78
Upvotes percentile
0.8963317384370016
Tags
—
Fetched at
Oct. 1, 2026, 1:01 a.m.
Updated at
Oct. 1, 2026, 1:01 a.m.

Description

Benchmarks how well Harness+models can create a Pac-Man game from a single prompt:“Create a Pac-Man game in a single HTML page”Each model gets one shot — no follow-up prompts or fixes.

Enrichment

Theme
Vertical
Horizontal
Function
Observability & eval
Audience
Developer
AI stance
AI-native
Project type
Hobby / open-source project
Normalized one-liner
pac-man game benchmark for evaluating ai models
Manually corrected
False

Could you build this?

Yes This is a static benchmark display site and evaluation runner script that prompts LLM APIs, tests if the resulting HTML contains valid Pac-Man code, and visualizes token/cost metrics.

Discussion

20 comments analyzed.

Concerns raised: Pac-Man is heavily represented in training data, One-shot benchmark is uninteresting and unrepresentative, Implementations lack subtle game mechanics and fidelity

Feature requests: Benchmark with novel variations not in training data, Test with Opus 5.5

Competitors

Other products that read as similar to this one — 13 launches clear the similarity bar, closest 8 shown.

Attention rank: #4 of 14 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).

Launched 281 days after the earliest competitor.

Other launches for this product

Same idea, different domain

Nobody's really built a observability & eval tool for Media & entertainment yet.