So OpenAI recently revealed a new pricing structure for its advanced AI agents, and the numbers are eye-opening. The company plans to introduce three distinct tiers for its AI services:

• Low-end agents at $2,000 per month, targeting high-income knowledge workers for tasks like data analysis and report generation.
• Mid-tier agents priced around $10,000 per month, designed specifically for software development tasks such as autonomous coding and debugging.
• High-end agents costing up to $20,000 per month, positioned as PhD-level research assistants capable of in-depth analysis and hypothesis testing.

These numbers are a lot higher than I would have imagined. We always think AI is a fraction of the cost compared to human labor, but this pricing reveals that perhaps things are pricier than they initially seem after running the numbers.

Consider the era when Lyft regularly offered $5 ride promos or when BlueApron consistently sent out $50 discount coupons to entice return users. Those deals were heavily subsidized by venture capital money to capture market share quickly. Could something similar be happening with AI?

What Happens When VC Subsidies Dry Up?

The current affordability of many AI services might be heavily subsidized by venture capital funding. Tech giants have been absorbing massive compute and R&D costs, artificially lowering prices for end-users. If these subsidies disappear—as they inevitably will—prices for enterprise-grade AI solutions could rise significantly. This raises critical questions: Has the price of AI been artificially low due to VC money? Are we witnessing an inflection point where the true costs become transparent?

The introduction of OpenAI's pricey new tiers suggests that we're approaching a more realistic pricing model for advanced AI capabilities—one that reflects genuine operational and computational expenses rather than artificially subsidized rates. The age of artificially cheap AI may be ending sooner than we think.