SAN FRANCISCO, Calif., April 25, 2025 — OpenAI is expanding access to a streamlined version of its deep research tool for ChatGPT, offering what it calls a “lightweight” model to ChatGPT Plus, Team and Pro users, the company announced Thursday.

The new tool is powered by the company’s o4-mini model, which OpenAI says enables broader usage due to its lower operational cost. Unlike the full-featured version of the deep research assistant, the lightweight model generates shorter responses, though OpenAI claims it preserves the depth and factual quality that users expect.

Free-tier ChatGPT users will begin receiving access starting Thursday, with support for enterprise and education customers expected next week. According to OpenAI, usage limits for the full deep research tool will remain in place, but queries will automatically transition to the lightweight version once those limits are reached.

“Responses will typically be shorter while maintaining the depth and quality you have come to expect,” OpenAI said in a statement on X. “Once limits for the original version of deep research are reached, queries automatically default to the lightweight version.”

The expansion comes amid a surge in generative AI platforms offering similar functionality. Google’s Gemini, Microsoft’s Copilot, and Elon Musk’s xAI-backed Grok have all recently introduced or enhanced their own research-focused chatbot features. These tools are often underpinned by so-called “reasoning models,” which are designed to break down complex problems, verify information, and simulate critical thinking — abilities considered essential for AI tasked with compiling research or summarizing information across sources.

OpenAI’s move reflects a broader shift in the AI sector toward balancing performance with accessibility. By using a lighter-weight model to reduce processing demands, the company can allow more users to engage with its research capabilities without sacrificing reliability, a critical consideration as demand for ChatGPT continues to grow globally.

With enterprise adoption on the rise — particularly in education, software development, and media sectors — OpenAI is positioning itself to compete not only on model performance but on usability at scale. While the full capabilities of the o4-mini model remain undisclosed, early users have reported that the lightweight version performs capably across general research topics, albeit with fewer citations and shorter summaries.

The company did not disclose whether the lightweight research tool will eventually support multimedia or proprietary document analysis, features currently being piloted in other versions of ChatGPT for enterprise clients.

As generative AI becomes increasingly embedded in workplace and academic settings, OpenAI’s incremental rollout strategy may help it gather user feedback and optimize performance ahead of a broader deployment. With competitive pressure mounting and user expectations rising, the introduction of a cost-effective research assistant represents a calculated move to consolidate market share while keeping infrastructure costs in check.