Frontier3DBench: 145-task benchmark tests coding agents on building interactive 3D worlds
_akhaliq · x · 2026-11-06
Einsia AI released Frontier3DBench, a unified benchmark evaluating coding agents on full-stack 3D creation: objects, scenes, and interactive open worlds.
Benchmark design
- 145 tasks across 9 families and 3 levels: L1 Objects (65), L2 Scenes (65), L3 Interactive open worlds (15)
- Native Blender and Unreal Engine 5 projects that can be opened, edited, and executed
- 11 frontier model-harness configurations, jointly scored on appearance, structure, and behavior
Leaderboard highlights (total = 30% L1 + 30% L2 + 40% L3)
- GPT-6.1 Sol (Codex harness) tops at 66.02
- GPT-6 Astra 63.85, Claude Opus 5.5 (Claude Code) 62.29
- DeepSeek V4.1 Flash 57.45, Kimi K3.1 Pro 55.96
- Scores drop sharply on L3 interactive worlds, making interaction and persistent state the main bottleneck
Project page, GitHub repo, and paper are all public.
Related event: Frontier3DBench Tests Coding Agents on 145 3D World-Building Tasks(2 posts)→
More from coding & agent
- Dev shows multi-agent crew auto-creating a Black Friday discount code end-to-end for $0.10 — pramodk73 · 2026-11-06
- Developer says Codex wrecked his server after a reboot request, costing $10,000+ — burny_tech · 2026-11-06
- Exo: Berkeley's agent runtime that edits its own harness, cutting model cost 35-61% — dair_ai · 2026-11-06
- Symbolic belief revision plus geometric projections lifts RAG reranking to 0.481 nDCG@10 — letandrewcook · 2026-11-06
- Free open-source AI Engineering from Scratch course teaches math-first — HowDevelop · 2026-11-06
- Charter: A Lean 4 DSL for Formally Verified Rules Governing AI Agents — QuixiAI · 2026-11-06