OpenAI price reduction GPT-4o: input price dropped to $2.5/ million tokens, a drop of more than 30%

One sentence conclusion

OpenAI lowers GPT-4o input price by 30% to $2.5/ million tokens, which is another important signal for the AI price war in 2026. For developers, monthly API costs can save hundreds to thousands of dollars; for entrepreneurs, the threshold for AI applications is further lowered.

What happened

In April 2026, OpenAI officially adjusted the API pricing of the GPT-4o model. The core change is that the input price has dropped from the original $3.5/ million tokens to $2.5/ million tokens, a decrease of 28.6% (nearly 30%). The output price remains unchanged for the time being, still $10.5/ million tokens.

This price cut is not an isolated incident. During the same period, OpenAI also reduced the price of the Pro package from $200/ months to $100/ months, a drop of 50%. These two adjustments together constitute OpenAI’s combination of punches to deal with market competition.

The background is that the AI ​​model price war will continue to heat up in 2026. Anthropic's Claude Opus 4.7 and Google's Gemini 2.0 Pro both put pressure on OpenAI in terms of price and performance. OpenAI needs to maintain market share through price cuts while paving the way for the upcoming release of GPT-6.

Direct impact on practitioners

Cost level

Take a typical medium-sized AI application as an example:

• Before Price Reduction: Processing 5 million input tokens per month at cost of $3.5 × 5 = $17.5
• After price reduction: The cost for the same usage is reduced to $2.5 × 5 = $12.5
• Monthly Savings: $5 (28.6% savings)

For heavy users (e.g. 10 million tokens per month):

• Monthly cost reduced from $35 to $25, saving $10
• Annual savings of up to $120

Efficiency/capability level

Falling prices mean:

1. Reduced Experimentation Cost: Developers can test new features more freely and try different prompt word strategies
2. Batch processing is more economical: Processing long documents and batch generating content are more economical
3. Edge Use Cases Become Viable: Features that were previously discarded as too costly can now be re-evaluated

Compliance/Risk Level

No direct impact. OpenAI’s terms of service and compliance requirements remain unchanged.

Core data comparison

Our judgment

**This price reduction is underestimated by the market. **

On the surface, it looks like a 30% price reduction, but the actual impact is far more than that. There are three reasons:

First, this is OpenAI’s fourth major price cut since 2023. Each price cut is accompanied by a wave of innovation in AI applications. $2.5/The price point of one million tokens makes more "small but beautiful" AI functions economically feasible.

Second, the price reduction coincides with competitors tightening their free policies. Google has just moved Gemini Pro out of the free tier, and OpenAI has a clear competitive intention to lower the price at this time. For teams that are evaluating different APIs, OpenAI’s cost-effectiveness is even more obvious.

Third, price reduction paves the way for the GPT-6 ecosystem. OpenAI needs to expand its developer base and accumulate users and data for next-generation models. The current price reduction is a strategic investment, not a simple promotion.

Those who benefit the most are:

• AI startups in the prototype stage
• SaaS tool that relies heavily on text processing
• Budget-sensitive areas such as education and content creation

Minimal impact is:

• Large enterprises that already use caching and batch APIs heavily
• Applications dominated by output tokens (the output price remains unchanged)

3 things you can do now

1. Re-evaluate the monthly budget: Log in to the OpenAI console, view the actual usage in the past 3 months, and calculate the new cost after the price reduction. If your monthly usage exceeds 1 million tokens, consider upgrading to the Pro plan for more quota.

2. Test features that were previously abandoned due to cost: Those features that were "if only there was more budget" can now be re-evaluated. Especially in scenarios where long documents need to be processed and summaries are generated in batches.

3. Compare competing product prices: Compare OpenAI’s new price with Anthropic Claude and Google Gemini. If the main usage scenario is text processing, OpenAI's cost performance may already be ahead.

Related tools (jump within the site)

This article involves tools: OpenAI API, GPT-4o