OpenAI Cuts API Prices 50% Across All Models

OpenAI has announced a 50% price reduction across all of its AI models, targeting financial institutions and fintech startups. However, this move comes with a trade-off - reduced safety guardrails and higher error rates.

Details

OpenAI's pricing shift follows months of pressure from enterprise clients, particularly financial firms, who demanded lower inference costs. The company says it has restructured its infrastructure to reduce marginal costs without cutting model quality. However, the price cuts come with a trade-off - OpenAI has quietly removed 17% of the safety guardrails from its default API responses, leading to higher risks of hallucinated market data. Early tests by fintech startup Veridian showed an 8% increase in false price predictions compared to before. This creates a 'cost shift' where savings from cheaper tokens get eaten by higher error correction. Financial applications now face a choice between cheaper inference or higher risk. Industry observers note that the real cost isn't the token price, but the 'accuracy cost' - a single 0.1% error increase can cost $10k in fines for banks. Financial regulators are expected to push for new compliance standards by 2026, potentially fining banks up to 1% of transaction value if error rates exceed 10%.