GroupTravelBench - Plan Validity: leaderboard
GroupTravelBench Plan Validity is the share of group itineraries produced by an LLM agent that pass nine deterministic checks for transport, lodging, scheduling and cost, over 650 multi-user travel planning tasks (GroupTravelBench, 2026). Higher is better.
Metric: Plan Validity (%): share of final itineraries that pass all nine deterministic feasibility checks (transport, lodging, scheduling, cost), over all 650 tasks, on GroupTravelBench, multi-user multi-turn group travel planning in a group-chat sandbox with 10 cached travel tools (DeepSeek-V4-Flash user simulator at temperature 0, agent temperature 0.7, mean of 3 trials per task); GPT-5.1 in instruct mode, the other models with reasoning enabled; higher is better. Source: arxiv.org. Saturation forecast: Around July 2028. 9 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Qwen 3.5 Plus | Thinking | 12 | #242 |
| 2 | DeepSeek V4 Pro | Thinking | 8 | #196 |
| 3 | Qwen 3.6 Max (Preview) | Thinking | 7 | |
| 4 | GroupTravelBench Qwen3-235B-A22B-Thinking (checkpoint unspecified) | 3 | ||
| 5 | GPT-5.1 | Non-reasoning | 1 | #240 |
| 6 | Qwen 3 4B Instruct | 1 | #1696 | |
| 7 | Qwen 3.5 Flash | Thinking | 1 | #441 |
| 8 | GroupTravelBench Qwen3-30B-A3B-Thinking (checkpoint unspecified) | 1 | ||
| 9 | GroupTravelBench Qwen3-4B-Thinking (checkpoint unspecified) | 0 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=grouptravelbench-plan-validity · How It Works · Data refreshed daily, snapshot 2026-11-21.