Trip+ - Response Mode Accuracy: leaderboard
Trip+ (2026) has tool-using LLM agents plan and revise minute-level travel itineraries over multi-turn conversations with profiled travelers, changing requests and environment disruptions, in a fixed travel sandbox. This board is response-mode accuracy: whether the agent plans, asks for clarification or reports no solution when the turn calls for it. Scores are on a 0 to 100 scale; higher is better.
Metric: Response-mode accuracy (0-100), a rate scaled by 100: whether the agent answers a turn with a plan, a clarification or a no-solution reply as the hidden turn state expects, averaged over all turns, on multi-turn personalized travel planning with traveler profiles, request changes and environment disruptions, the same OpenAI-compatible function-calling scaffold over a fixed travel sandbox for every model, temperature 0 where supported; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemma 4 31B | 90.35 | #394 |
| 2 | Hy3-preview | 89.3 | #316 |
| 3 | GLM-5.1 | 87.02 | #179 |
| 4 | Gemini 3 Flash (Preview) | 86.67 | #144 |
| 5 | Gemini 3.1 Pro (Preview) | 85.44 | #94 |
| 6 | Qwen 3.5 27B (Non-reasoning) | 84.91 | #359 (Qwen 3.5 27B) |
| 7 | Kimi K2.6 | 84.39 | #178 |
| 8 | Gemma 4 26B A4B | 83.86 | #615 |
| 9 | Seed 2.0 Pro | 82.46 | #211 |
| 10 | DeepSeek V4 Pro | 79.12 | #188 |
| 11 | GPT-5.4 Mini | 78.07 | #389 |
| 12 | MiniMax-M2.7 | 77.72 | #518 |
| 13 | DeepSeek V3.2 | 77.02 | #406 |
| 14 | GPT-5.4 | 75.44 | #123 |
| 15 | Qwen 3.6 35B A3B (Non-reasoning) | 69.12 | #462 (Qwen 3.6 35B A3B) |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=trip-plus-response-mode-accuracy · How It Works · Data refreshed daily, snapshot 2026-11-13.