TravelEval - Experience Diversity: leaderboard
TravelEval (KDD 2026) asks an LLM for a complete multi-day travel plan and simulates it in a sandbox of 10 Chinese cities with real transport, hotel and attraction data. Scores are from Direct prompting over 1,150 queries. The board is the normalized entropy of the experience types a plan covers, on a 0 to 100 scale; higher is better.
Metric: Experience Diversity Index (EDI): normalized Shannon entropy of the experience types visited, times 100 (0-100); Direct prompting (one pass, no tools), plans simulated in the TravelEval sandbox of 10 Chinese cities (real rail, flight, hotel and attraction data, queuing-time model, road-network distances) over 1,150 queries; higher is better. Source: arxiv.org. Saturation forecast: Around March 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Sonnet 4.5 | 62.35 | #250 |
| 2 | Gemini 2.0 Flash | 61.76 | #704 |
| 3 | DeepSeek V3.1 (Non-reasoning) | 61.52 | #497 (DeepSeek V3.1) |
| 4 | GPT-5 Chat | 61.12 | #444 |
| 5 | GPT-4o Mini (2024-07-18) | 56.27 | #1510 |
| 6 | GPT-4o (2024-11-20) | 56.27 | #866 |
| 7 | Qwen 3 8B | 52.71 | #1830 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=traveleval-experience-diversity · How It Works · Data refreshed daily, snapshot 2026-11-13.