IndicBankBench - Multi-Tool Agentic: leaderboard

Metric: Strict pass^3 (%; 112 cases passed out of the 112 cases of the general multi-tool agentic 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
1Claude Opus 5 (Max)42.86
2GLM-5.3 Flash (High)39.29
3Gemini 3.8 Flash38.39
4Grok 4.6 (High)37.5
5DeepSeek V4 Pro (0813) (Max)37.5
6DeepSeek V4 Flash (0731) (Max)33.04
7Gemma 4 31B (IT) (Thinking)31.25
8MiniMax-M324.11
9gemma-4-26B-A4B-it (Thinking)21.43
10Gemma 4 E4B (IT) (Thinking)19.64

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