OpenAI has released a new “lightweight” version of its ChatGPT deep research tool.

This model, powered by OpenAI’s o4-mini, is accessible to both free users and those on various paid subscription tiers.

This lightweight version offers shorter replies while keeping the quality of deep research mode.

Free users get up to five tasks per month, while Team and Plus subscribers receive 25, and Pro users get 250.

Enterprise and Education users will gain access next week with similar usage limits.

The tool is cheaper to run than the standard version, helping OpenAI reduce processing costs.

🔗 Source: The Verge

🧠 Food for thought

1️⃣ Cost efficiency drives AI product tiering amid soaring development expenses

OpenAI’s introduction of a “lightweight” deep research model reflects an industry trend of managing the high costs associated with advanced AI systems.

Training leading AI models can cost upwards of $192 million, as seen with models like Google’s Gemini, creating significant operational expenses for companies1.

This cost pressure is driving a market-wide shift toward more efficient AI solutions, with companies developing specialized systems that balance capabilities with operational expenses2.

By offering a cheaper-to-operate lightweight version, OpenAI aligns with the cost-efficiency trend identified in the Stanford AI Index, where companies are seeking ways to maintain performance while reducing the substantial computational resources required3.

The company’s acknowledgment that it spends “tens of millions of dollars” on processing simple user interactions highlights the scale of operational costs that AI providers are working to optimize.

2️⃣ AI tool marketplaces evolve with standardized tiered access models

OpenAI’s approach of offering limited functionality to free users while reserving full capabilities for paying customers reflects a common pattern in the AI chatbot marketplace.

Recent OpenAI developments