DeepPlanningAgents

DeepPlanning evaluates LLMs on complex planning tasks such as multi-day travel itineraries and multi-product shopping. Agents must actively gather information, reason over local constraints, and optimize global time and budget constraints that can be verified.

Ranking Sources

#ModelScore

1

Qwen3.7-PlusAlibaba

62.3%

2

Claude Opus 4.6Anthropic

58.9%

3

DeepSeek-V4-ProDeepSeek

44.6%

4

Kimi-K2.6Moonshot AI

42.3%

5

Qwen3.6-PlusAlibaba

41.5%

6

Qwen3.5-397B-A17BAlibaba

37.6%

7

GLM-5.1Zhipu AI

34.1%

8

Claude Opus 4.5Anthropic

33.9%

9

Qwen3.6-35B-A3BAlibaba

25.9%

10

Qwen3.5-35B-A3BAlibaba

22.8%

12 models total

Back to benchmarks