LiveBench Logic With Navigation: leaderboard
Each item pairs a logic problem with movement in a 2D space: the model solves the logic problem, navigates, and reports the final x and y coordinates inside solution tags. LiveBench compares that coordinate pair with the ground truth, so scoring is automatic, one point or zero, and uses no judge model. This board is that one task's average score on the current LiveBench release, higher is better.
Metric: Score. Source: livebench.ai. Saturation forecast: Around January 2027. 71 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra (Max) | 88 | #6 (GPT-6 Astra) |
| 2 | Muse Spark 1.2 (xHigh) | 88 | #62 (Muse Spark 1.2) |
| 3 | Claude Opus 5 (Max) | 86 | #12 (Claude Opus 5) |
| 4 | GPT-6.1 Sol (Max) | 86 | #7 (GPT-6.1 Sol) |
| 5 | Claude Fable 5.1 (Max) | 86 | #9 (Claude Fable 5.1) |
| 6 | Claude Sonnet 5 (xHigh) | 86 | #57 (Claude Sonnet 5) |
| 7 | Claude Opus 5.5 (Max) | 86 | #3 (Claude Opus 5.5) |
| 8 | Qwen 3.7 Max | 84 | #75 |
| 9 | GPT-5.6 Sol (Max) | 82 | #19 (GPT-5.6 Sol) |
| 10 | Gemini 3.8 Flash (High) | 82 | #21 (Gemini 3.8 Flash) |
| 11 | Grok 4.6 | 82 | #26 |
| 12 | GPT-6.1 Sol (xHigh) | 82 | #7 (GPT-6.1 Sol) |
| 13 | Claude Sonnet 5.5 (Max) | 82 | #11 (Claude Sonnet 5.5) |
| 14 | GLM-5.2 | 80 | #73 |
| 15 | Kimi K3 | 80 | #30 |
No result here: #8 Gemini 4 Argon, #10 Muse Wonder Full.
Interactive version: theaggregate.ai/benchmark?slug=livebench-logic-with-navigation · How It Works · Data refreshed daily, snapshot 2026-10-20.