PulseAugur
EN
LIVE 03:55:55
한국어(KO) 동일한 한 가지 프롬프트로 두 AI 코딩 에이전트가 브라우저용 오픈월드 드라이빙 샌드박스를 자율적으로 기획·구현·테스트·반복해 완성한 결과를 비교. Claude Fable 5.1은 'Sunset City'(약 6h15m, 토큰 ~4.58M, 다수 인력) 를, Codex Astra 6는

AI coding agents Claude Fable 5.1 and Codex Astra 6 compete in sandbox game development

By PulseAugur Editorial · [1 sources] ·

Two AI coding agents, Claude Fable 5.1 and Codex Astra 6, were tasked with autonomously developing a browser-based open-world driving sandbox using a single prompt. Claude Fable 5.1 produced 'Sunset City' in approximately 6 hours and 15 minutes, utilizing multiple agents and around 4.58 million tokens. Codex Astra 6, on the other hand, created 'Palm Bay' in about 5 hours and 16 minutes, using a single agent and approximately 7.23 million tokens. Both games feature WASD/E/Space controls and a low-poly art style, with an emphasis on automated review and iterative improvement. AI

IMPACT Demonstrates the differing capabilities and resource utilization of AI agents in complex creative tasks like game development.

RANK_REASON Comparison of two AI coding agents' output on a specific task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on Mastodon — fosstodon.org →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

AI coding agents Claude Fable 5.1 and Codex Astra 6 compete in sandbox game development

How we ranked this

Signal score
17 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
Comparison of two AI coding agents' output on a specific task. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
product, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

Full methodology in our editorial standards.

COVERAGE [1]

  1. Mastodon — fosstodon.org TIER_1 한국어(KO) · [email protected] ·

    Comparison of results where two AI coding agents autonomously planned, implemented, tested, and iterated on an open-world driving sandbox for browsers using the same single prompt. Claude Fable 5.1 completed 'Sunset City' (approx. 6h15m, tokens ~4.58M, multiple personnel), and Codex Astra 6

    동일한 한 가지 프롬프트로 두 AI 코딩 에이전트가 브라우저용 오픈월드 드라이빙 샌드박스를 자율적으로 기획·구현·테스트·반복해 완성한 결과를 비교. Claude Fable 5.1은 'Sunset City'(약 6h15m, 토큰 ~4.58M, 다수 인력) 를, Codex Astra 6는 'Palm Bay'(약 5h16m, 토큰 ~7.23M, 단일 에이전트)를 제작. 게임은 WASD/E/스페이스 조작, 저폴리 스타일과 자동 리뷰·반복 개선을 핵심으로 한다. https:// nitzan.games/On…

    LINKS nitzan.games/OneShotGTA.html

RELATED ENTITIES

RELATED TOPICS