Bengaluru: India has emerged as one of the fastest-growing global markets for
OpenAI’s coding platform, Codex. Adoption has surged nearly 27 times since the beginning of the year, while daily interactions rose more than 20 times by the end of April, according to Thomas Jeng, head of startups for Asia-Pacific (APAC) at OpenAI.
Jeng said India is now among Codex’s top five markets globally by user count, driven by a rapidly growing base of developers and startups using the platform to automate software development and knowledge workflows.
“Indian engineers comprise some of the most sophisticated builders in the world,” Jeng told TOI. “They’re using Codex in ways that we may not have anticipated. Teams are shipping products much faster, moving from what may have taken years earlier to just a few months before production deployment.”
Codex, which initially started as a coding assistant, has evolved into a broader “agentic workflow” platform capable of automating long-running tasks, spawning multiple workstreams and coordinating sub-agents. According to Jeng, adoption accelerated significantly after the launch of the Codex application, which provided developers with a more user-friendly interface to access advanced AI capabilities.
The comments come amid intensifying competition in the AI coding assistant market, where Anthropic’s Claude Code has also gained traction among developers. While acknowledging that competing products entered the category earlier, Jeng said the adoption gap is narrowing rapidly.
“The feedback we consistently hear is that Codex behaves more like a senior engineer or solutions architect,” he said. “It performs more planning, reviews codebases more thoroughly and produces higher-quality outputs.”
The data also points to a widening role for Codex in India beyond software engineering. More than a quarter of Codex requests are now for non-coding tasks, reflecting growing adoption across research, documentation, workflow management and information synthesis. Users are increasingly relying on the platform to automate repetitive daily tasks, organise communication and streamline operational workflows.
OpenAI is also expanding its footprint in India through startup partnerships, hackathons, enterprise collaborations and venture capital alliances. Jeng said the company now works with dozens of VC firms in India through startup credit programmes that offer up to $50,000 in OpenAI credits globally.
“We’ve already distributed millions of dollars’ worth of credits in India,” he said, adding that the partnerships are designed to create a “win-win-win” for startups, investors and OpenAI.
The company is also deepening ties with enterprises and IT services firms, including TCS, Infosys and Razorpay, to expand deployment of Codex and other OpenAI technologies.
Despite concerns around rising AI infrastructure and token costs, Jeng argued that AI economics are improving at the task level even as overall spending rises.
“If companies give engineers unlimited tokens, costs will obviously escalate,” he said. “But newer models can actually reduce the cost per task compared with earlier systems.”
He added that OpenAI does not see Codex as a replacement for human developers but rather as a productivity amplifier.
“The most proficient developers using these tools are often the most productive developers overall,” Jeng said. “Enterprises will ultimately find the right balance between human developers and AI tooling depending on their use cases.”