IndicBankBench - Irrelevant RAG: leaderboard

Metric: Strict pass^3 (%; 24 cases passed out of the 24 cases of the irrelevant-retrieval axis; each case run three times at temperature 0.7 with the stated reasoning modes; a case passes only when all three runs pass the deterministic safety and tool-use gates and the GLM-5.2 response judge; scripted Indian retail-banking conversations over deterministic mock tools). Source: arxiv.org. Saturation forecast: Around 2029. 11 models tracked.

Top models

#ModelScore
1Grok 4.6 (High)75
2Gemini 3.8 Flash70.83
3DeepSeek V4 Pro (0813) (Max)70.83
4DeepSeek V4 Flash (0731) (Max)70.83
5Gemma 4 31B (IT) (Thinking)70.83
6gemma-4-26B-A4B-it (Thinking)70.83
7Gemma 4 E4B (IT) (Thinking)66.67
8MiniMax-M362.5
9GLM-5.3 Flash (High)62.5
10Claude Opus 5 (Max)58.33

Interactive version: theaggregate.ai/benchmark?slug=indicbankbench-irrelevant-rag · How It Works · Data refreshed daily, snapshot 2026-09-26.