How much does the OpenAI API cost per token?

OpenAI cost per token ranges from $0.0000002 (GPT-5.4 Nano input) to $0.00018 (GPT-5.4 Pro output). To convert from OpenAI’s published per-million-token rates, divide by 1,000,000. Most production workloads fall between $0.0000002 and $0.000015 per token depending on the model. You can use the formula in our calculator section above to estimate your own OpenAI API cost per token for any model.

Is the OpenAI API free?

OpenAI offers an OpenAI API pricing free tier with limited rate limits and credits for new accounts. Beyond the free credits, all API usage is billed per token. There is no free unlimited tier. The OpenAI API key cost is $0 to generate, but every API call after your free credits are used incurs token charges. ChatGPT Free provides limited access to GPT-5.4 Mini, but this is the consumer product, not the API.

How much does it cost to run a chatbot on the OpenAI API?

A chatbot handling 10,000 conversations per day on GPT-5.4 Mini costs roughly $518 per month. The exact OpenAI API cost depends on average conversation length, model choice, and caching efficiency. Switching from Standard to Mini reduces costs by about 70% for most support use cases.

What is the cheapest OpenAI model?

GPT-5.4 Nano is the cheapest production model at $0.20 per million input tokens and $1.25 per million output tokens. When comparing GPT-5 pricing across the full lineup, Nano costs 12x less than Standard on input. GPT-OSS-20b, an open-source model, is available at $0.03 per million tokens but has limited capabilities.

How does OpenAI API pricing compare to Anthropic Claude?

GPT-5.4 Mini ($0.75 input / $4.50 output) is roughly 25% cheaper on input than Claude Haiku 4.5 ($1.00 input / $5.00 output) and 10% cheaper on output. GPT-5.4 Standard ($2.50 / $15.00) is slightly cheaper on input than Claude Sonnet 4.6 ($3.00 / $15.00) and matches on output. At the flagship tier, Claude Opus 4.6 costs $5.00 / $25.00 per million tokens. Comparing the full OpenAI API price lineup to Claude, OpenAI is generally cheaper at the lower tiers.

What is the OpenAI Batch API and how much does it save?

The Batch API processes requests asynchronously within 24 hours at 50% of standard token prices. GPT-5.4 Standard drops from $2.50/$15.00 to $1.25/$7.50 per million tokens. It is ideal for workloads that do not need real-time responses, like nightly data processing, bulk content generation, or large-scale classification.

Is there an OpenAI cost calculator?

OpenAI does not offer a dedicated OpenAI cost calculator, but you can estimate costs with a simple formula: (input tokens / 1,000,000 x input price) + (output tokens / 1,000,000 x output price). OpenAI’s usage dashboard also tracks token consumption in real time and projects monthly spend.

How do Azure OpenAI costs compare to the direct API?

Azure OpenAI pay-as-you-go rates are generally comparable to direct OpenAI API pricing. Azure also offers provisioned throughput units (PTUs) for reserved capacity at volume discounts. The main reason to choose Azure is compliance, data residency, or integration with existing Azure infrastructure, not price.

CloudZero Blog

Artificial Intelligence (AI)

The incorporation of AI into our daily work and lives is one of the most significant paradigm shifts to hit us in decades. Here, we discuss AI in general, and particularly how it relates to cloud spend.