Portal2Bench releases AI agent puzzle-solving benchmark
Understand this story
Portal2Bench has released a benchmark that has AI agents independently solve puzzles in Portal 2 test chambers. Each chamber is a fresh, independent run, and an agent must reach the next map within 300 planning steps and at a cost of no more than $30 to pass; costs are calculated at API list prices. Among the published results, GPT-6.1 Sol and GPT-6 Astra each cleared 20 of 26 chambers, Opus 5.5 cleared 11, Sonnet 5.5 cleared 6, and GPT-6 Luna cleared 0. The testers say all models ran at maximum intensity, with Gemini 4 Argon running at high intensity; due to compute and usage quota limits, most chambers are still running.
AI overview based on source reports · Updated 13 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: Portal2Bench – Can AI agents solve Portal 2?
Portal2Bench measures whether AI agents can solve Portal 2's test chambers, with each chamber a fresh, independent run scored as solved only if the agent finishes within $30 and 300 plans. In the reported results, GPT-6.1 Sol (Codex) and GPT-6 Astra (Codex) each solve 20 of 26 chambers, GPT-6.1 Terra 19, Opus 5.5 11, Sonnet 5.5 6, GPT-6 Luna 0, and Fable 5.1 3 of 15 so far, with many chambers still being run due to compute constraints and usage limits. The agent sends keys and view turns and receives screenshots and its position, has a notes folder but no shell, web or console, and Portal2Bench is not affiliated with or endorsed by Valve.
Activity for this story
- Current comparable activity
- 7
- Peak comparable activity
- 10Nov 14 09:00
- Change in the past 24 hours
- –
Trends compare the same fully observed participants, which may cover fewer participants than current activity. Move the pointer or tap the chart to inspect hourly activity; use the left and right arrow keys with a keyboard.