PlanBench-XL: leaderboard
LLM agents increasingly operate in large tool ecosystems, where real-world tasks require discovering relevant tools, inferring implicit sub-goals, and adapting to dynamic environments over long horizons. However, existing benchmarks rarely evaluate planning under retrieval-limited tool visibility. To address this gap, we introduce PlanBench-XL, an interactive benchmark of 327 retail tasks over 1,6
Metric: Accuracy (%) (self-reported). Source: benchmarklist.com. Saturation forecast: Around April 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 77.06 | #58 |
| 2 | DeepSeek V4 Flash | 63.08 | #137 |
| 3 | Gemini 3.5 Flash | 52.19 | #60 |
| 4 | GPT-5.4 | 51.9 | #79 |
| 5 | Llama 3.3 70B Instruct | 18.96 | #534 |
| 6 | GPT-5.4 Mini | 3.07 | #212 |
| 7 | Qwen 3 32B | 2.75 | #443 |
| 8 | Qwen 3 14B | 0.92 | #537 |
| 9 | Llama 3.1 8B Instruct | 0 | #1035 |
| 10 | Qwen 3 8B | 0 | #674 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=planbench-xl · How It Works · Data refreshed daily, snapshot 2026-10-20.