The Race to the Bottom Continues
In what many analysts are calling the most aggressive pricing move of 2026, OpenAI has announced a sweeping 30% reduction in API costs for its flagship GPT-4o model. The new pricing takes effect immediately, dropping input token costs from $2.50 to $1.75 per million tokens and output costs from $10.00 to $7.00 per million tokens.
This move comes just weeks after Google reduced Gemini 2.5 Flash pricing and Anthropic introduced more competitive tiers for Claude. The AI pricing landscape has shifted dramatically over the past 12 months, with costs dropping an average of 40-60% across all major providers.
What This Means for Developers
For teams running production workloads, this translates to immediate savings. A typical SaaS application processing 10 million tokens per day would see monthly costs drop from roughly $3,750 to $2,625 — a saving of over $1,100 per month without changing a single line of code.
The reduction also narrows the gap between GPT-4o and smaller models like GPT-4o mini. For many use cases, the quality-to-cost ratio of the full-size model now makes it viable even for high-volume applications.
Industry Response
Competitors are expected to respond quickly. Anthropic and Google typically match major pricing moves within 2-4 weeks. DeepSeek, already the cost leader for reasoning tasks, may not need to respond directly but could use this as an opportunity to highlight its value proposition.
Use our cost calculator to see exactly how these new prices affect your specific workload.