SAGE (Service Agent) - Logical Compliance: leaderboard

SAGE logical compliance: whether a customer-service agent classifies the user state, plans the procedure path and takes the action that the dialogue graph and a rule engine prescribe. 0-100 scale, higher is better. It is unrelated to the SAGE robustness benchmark.

Metric: Logical Compliance Score (0-100): 0.4 classification accuracy + 0.4 path correctness + 0.2 action correctness against the graph, averaged over the six scenarios, over SAGE customer-service dialogues: standard operating procedures formalized as dialogue graphs, an adversarial simulated user with varied intents and personas, turns 1, 5, 10, 15 and the final turn scored; three unnamed judge LLMs (majority vote and mean) plus a deterministic rule engine produce the reference; closed models through their APIs, open models served with vLLM; 0-100 scale; higher is better. Source: arxiv.org. Saturation forecast: Around March 2027. 27 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Opus 4.572.1#165
2DeepSeek V3.272.08#425
3Gemini 3 Pro (Preview)71.27#123
4Claude Sonnet 4.570.1#267
5DeepSeek R170#484
6Kimi K2.569.16#272
7GLM-4.768.68#383
8GPT-4.168.31#506
9Gemini 3 Flash (Preview)68.16#156
10DeepSeek V367.9#736
11Qwen Max (SAGE checkpoint unspecified)67.57
12MiniMax-M2.166.59#582
13Qwen 2.5 32B Instruct66.49#1253
14Llama 3.3 70B Instruct66.06#1394
15Qwen 3 32B65.97#983

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=sage-service-agent-logical-compliance · How It Works · Data refreshed daily, snapshot 2026-11-22.