Meet Artham: India’s First Stock-Savvy AI Model Built Just for Indian Markets
So, Indian fintech just dropped its own AI stock-market oracle, and it’s not even pretending to be neutral. Artham, the shiny new 7-billion-parameter Small Language Model from Raise (Dhan’s parent), isn’t just another global LLM with a desi accent—it’s trained exclusively on Indian financial data, regulations, and market jargon. Translation: it doesn’t just know what a stock is; it knows what a stock does when the RBI sneezes. This isn’t AI for the sake of buzzwords; it’s AI that’s supposed to connect the dots between corporate actions, macro shifts, and your portfolio pain, all while staying firmly within India’s data sovereignty sandbox. If this works, retail investors might finally get something closer to institutional-grade insight, minus the six-figure price tag.
Artham is positioned as India’s first SLM built specifically for Indian capital markets, already powering platforms like Dhan, Fuzz, and ScanX. It’s designed to deliver contextual, source-backed insights, generate real-time financial news at scale, and support secure, real-time market data access via native tool calling. Developed over nine months by Raise AI, the model emphasizes compliance, data security, and local relevance, running on SOC 2 Type II compliant infrastructure hosted entirely in India. By launching at AWS re:Invent 2025, Raise is signaling that India isn’t just consuming global AI—it’s now building sovereign, domain-specific models that could reshape how Indian investors, advisors, and media consume financial intelligence.
