Habr: Comparing AI Models While Building a Battle City Remake (2026)
Habr has published an updated comparison of AI models building a remake of the classic game Battle City (1985). The first comparison of different models on this task appeared six months ago. The experiment was repeated to show how fast the AI world moves. The results trace how the capabilities of models from OpenAI (GPT), Google (Gemini), Anthropic (Claude) and others have evolved in practical game development.
AI-processed from Habr AI; edited by Hamidun News
The Russian-speaking developer community has rerun an experiment comparing current AI models on a practical task: building a video game. It serves as an indicator of real progress in the capabilities of AI systems.
The original experiment
Six months ago, Habr published a first article comparing AI models on building a remake of the classic game Battle City. It tested the leading models of the time.
What changed in six months
- New model versions have shipped (GPT-5.x, Gemini 3+, Claude 3.5+ and others)
- Code generation has improved
- Most models now offer larger context windows
- Specialized coding models have appeared
Practical value
Comparing models on a real task is more informative than public benchmarks, because it shows:
- How well developers can work with unfamiliar code
- The ability to improve a solution iteratively
- Error handling and debugging
- The quality of the final working code
What this means
Repeating the experiment after a short interval (6 months) shows how fast the AI innovation cycle has become. For developers, this means the best tool keeps changing, and earlier recommendations may already be out of date. It also shows the practical relevance of AI for game development and other technical fields.
- Nov 1, 2026OpenAI Rewrites GPT Behavior Rules: Five Values and a Chain of Command
- Nov 1, 2026Google admits AI Mode cuts search revenue, shares fall 15%
- Oct 31, 2026Author puts his script before an AI jury: the verdict cost 850 thousand tokens
- Oct 31, 2026Anthropic: Fold Subagents Back Into a Single Loop — One Customer's Tasks Ran Twice as Fast
Key Questions Answered
What are the key details on The original experiment?
Six months ago, Habr published a first article comparing AI models on building a remake of the classic game Battle City. It tested the leading models of the time.
What changed in six months?
New model versions have shipped (GPT-5.x, Gemini 3+, Claude 3.5+ and others) Code generation has improved Most models now offer larger context windows Specialized coding models have appeared
What are the key details on Practical value?
Comparing models on a real task is more informative than public benchmarks, because it shows: How well developers can work with unfamiliar code The ability to improve a solution iteratively Error handling and debugging The quality of the final working code
Build & Ship Production AI Agents Without Boilerplate Code
Start your free 72-hour trial at Hamidun Academy (academy.hamidun.com) — interactive lessons on Claude Code & Cursor, live Tech Radar, and production agent templates.
The AI world, distilled — once a week
Seven stories that actually mattered, hand-picked. No noise, no reposts, no press releases.
Done! Check your inbox for a confirmation.