Updated daily at 05:00 UTC
Benchmark
Furniture Assembly
Measures: Spatial reasoning about physical construction: given a photo of a partly built IKEA product and its manual, say whether the build is correct and, if not, which step went wrong.
- Maintainer
- Epoch AI
- Introduced
- 2026
- Status
- Active
- Saturation
- Headroom
- Typical provenance
- Independent
- In scoring set
- Yes
Leaderboard
Every result we hold, whoever ran it. Only independent and third-party rows count towards a score.
| Model | Score | Provenance | Source | Evaluated |
|---|---|---|---|---|
| Qwen3.8-Max Alibaba Group · Xhigh effort | 20.0% | Independent | Epoch AI | 10 Sep 2026 |
| Grok 4.7 SpaceXAI (xAI) · Xhigh effort | 20.8% | Independent | Epoch AI | 24 Sep 2026 |
| Kimi K2.6 | 21.7% | Independent | Epoch AI | 11 Sep 2026 |
| Grok 4.5 SpaceXAI (xAI) · High effort | 22.5% | Independent | Epoch AI | 24 Sep 2026 |
| Gemini 3.6 Flash Google DeepMind · High effort | 23.3% | Independent | Epoch AI | 10 Sep 2026 |
| Gemini 3.1 Pro Google DeepMind · High effort | 26.7% | Independent | Epoch AI | 10 Sep 2026 |
| Gemini 3.7 Flash Google DeepMind · High effort | 26.7% | Independent | Epoch AI | 10 Sep 2026 |
| Claude Opus 4.5 Anthropic · 64K effort | 28.3% | Independent | Epoch AI | 10 Sep 2026 |
| Claude Opus 4.6 Anthropic · Max effort | 28.3% | Independent | Epoch AI | 10 Sep 2026 |
| Gemini 3.8 Flash Google DeepMind · High effort | 31.7% | Independent | Epoch AI | 10 Sep 2026 |
| Claude Opus 4.7 Anthropic · Max effort | 33.3% | Independent | Epoch AI | 10 Sep 2026 |
| Kimi K3 Moonshot AI · Max effort | 34.2% | Independent | Epoch AI | 11 Sep 2026 |
| Claude Fable 5 Anthropic · Max effort | 35.8% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-5.4 OpenAI · Xhigh effort | 37.5% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-5.2 OpenAI · Xhigh effort | 38.3% | Independent | Epoch AI | 10 Sep 2026 |
| Grok 4.6 SpaceXAI (xAI) · Xhigh effort | 40.0% | Independent | Epoch AI | 24 Sep 2026 |
| Claude Opus 4.8 Anthropic · Max effort | 42.5% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-5.6 Luna OpenAI · Max effort | 42.5% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-5.5 OpenAI · Xhigh effort | 44.2% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-6 Luna OpenAI · Max effort | 44.2% | Independent | Epoch AI | 28 Sep 2026 |
| GPT-5.6 Terra OpenAI · Max effort | 54.2% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-5.6 Sol OpenAI · Max effort | 56.7% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-6 Sol OpenAI · Max effort | 58.3% | Independent | Epoch AI | 23 Sep 2026 |
| Claude Opus 5 Anthropic · Max effort | 60.8% | Independent | Epoch AI | 10 Sep 2026 |
| Claude Fable 5.1 Anthropic · Max effort | 70.0% | Independent | Epoch AI | 10 Sep 2026 |
| GPT-6 Astra OpenAI · Max effort | 80.0% | Independent | Epoch AI | 10 Sep 2026 |
| Claude Opus 5.5 Anthropic · Max effort | 83.3% | Independent | Epoch AI | 22 Sep 2026 |
- Epoch AI, 'Capabilities & benchmarking', epoch.ai/benchmarks, CC BY 4.0.