OpenAI announced two major updates recently that underscore its twin priorities: Pushing the technical frontier of AI while making it culturally relevant for diverse global audiences. The company introduced GPT-5.1, its most advanced model yet, and IndQA, a benchmark designed to evaluate AI systems on Indian languages and cultural context.
GPT-5.1, brings two new interaction modes: Instant, designed for fast, conversational responses, and Thinking, which adapts reasoning for complex tasks such as coding, mathematics, and multi-step problem solving. The upgrade also improves instruction-following and tone customisation, allowing users to choose styles like Professional, Friendly, Candid, or even Quirky.
Paid users will get early access, with a broader rollout planned soon. OpenAI says GPT-5.1 significantly improves adaptive reasoning, outperforming previous versions on coding and math benchmarks, while making conversations feel more natural.
“AI is ushering in an entirely new era. It’s the most transformative technology we will experience in our lifetimes,” said Srinivas Narayanan, CTO of B2B Applications at OpenAI, at a recent press conference in Bengaluru. “We want OpenAI to be the best place for people to build. We have the best models in the world, and we also want to provide tools on top of those models to help people build truly remarkable applications.”
Read More
The company is positioning GPT-5.1 not just as a technical upgrade but as part of a broader ecosystem play. Developers can now build apps directly inside ChatGPT using the Apps SDK, and OpenAI is rolling out new agent-building tools to make AI integration easier across workflows.
Also Read: Quantum Dreams and AI realities: Sundar Pichai’s next big leap
IndQA: Making AI Understand India
While GPT-5.1 sets the pace globally, OpenAI’s IndQA initiative signals its commitment to India—a market that is both strategically important and culturally complex. IndQA is a benchmark built with 261 domain experts across 12 languages and 10 cultural domains, including literature, food, history, and everyday life. It comprises 2,278 questions, each natively written in Indian languages rather than translated, and uses a rubric-based evaluation to capture nuance. “India is home to incredible innovators… it’s been inspiring to see the innovation coming out of India, and we want to partner more closely with that community,” Narayanan said. “With IndQA, we’ve partnered with more than 250 experts to create a curated dataset that captures India’s cultural and historical nuances. Our hope is to use this as a playbook for other countries too.”Narayanan emphasised that AI cannot remain Western-centric if it aims to serve all of humanity. “India has such a diverse set of people, speaking so many different languages, and with varied cultures and a deep history. AI needs to understand all of that. It’s not just about building products from Silicon Valley that only understand the Western world,” he said.
OpenAI also introduced ChatGPT Go in India as part of its push for affordability and accessibility. The plan is free for a year, giving users access to premium features at no cost—a move aimed at expanding adoption among a wider audience. When asked about OpenAI’s roadmap for India, Narayanan outlined a multi-pronged approach: “India is our second-largest consumer market and the fastest-growing. We want ChatGPT to be available across all audiences. ChatGPT Go was the first time we created an SKU tailored for a new market to make AI more affordable. We’re using that as a template and have now launched it in 90 countries worldwide.”
On the developer side, OpenAI is betting big on agents: “You’ll see GPT-5, GPT-5.1, and beyond, helping push the frontiers of what individuals and companies can do. We’re building an entire suite of tools—Agent SDK, Agent Builder—to help developers create more agentic applications. Our goal is to help companies embrace this agentic future.”
OpenAI’s moves come amid intensifying competition with Google, Microsoft, and Meta, all of whom are investing heavily in AI infrastructure and developer tools. For OpenAI, India is not just a growth market but a proving ground for localisation strategies that could be replicated globally.
“At OpenAI, we care deeply about making AI useful for all of humanity. That means making it truly relevant for diverse cultures and languages,” Narayanan said. “Ultimately, it’s consistent with our vision to help meet developers where they are and help them solve problems for the local market.”
IndQA is expected to become a template for similar benchmarks in other countries, addressing a critical gap in AI evaluation: Cultural and linguistic depth. As Narayanan put it, “Our hope is that we take this as a playbook and apply it in other countries as well.”