GraphARC - Output Graph Node Count: leaderboard
Metric: Accuracy (%) of answers to questions about the number of nodes of the output graph that the inferred transformation produces (the model never sees that output), GraphARC few-shot graph-transformation tasks (21 transformations; the model sees a few input-output graph pairs and a test input graph of 5 to 15 nodes encoded as an adjacency or incidence list), answer compared with the value computed from the true output graph; the OpenAI reasoning models ran one system prompt at medium reasoning effort, the other models four system-prompt variants; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O3 Mini | 83 |
| 2 | O4 Mini | 69 |
| 3 | O1 Mini | 63 |
| 4 | Qwen 3 8B | 62 |
| 5 | GPT-5 | 61 |
| 6 | GPT-4.1 Nano | 49 |
| 7 | Qwen 3 32B | 49 |
| 8 | Qwen 3 14B | 39 |
| 9 | Qwen 3 1.7B | 39 |
| 10 | Llama 3.1 8B | 32 |
| 11 | Qwen 3 4B | 30 |
| 12 | Mistral-7B-v0.2 | 19 |
| 13 | OLMo 2 7B | 19 |
| 14 | DeepSeek R1 Distill Llama 8B | 15 |
| 15 | Llama 3 8B | 12 |
Interactive version: theaggregate.ai/benchmark?slug=grapharc-output-graph-node-count · How It Works · Data refreshed daily, snapshot 2026-10-07.