GPT-5 is available in both ChatGPT and the OpenAI API, with access and pricing varying by plan and model size. In ChatGPT, all users—including those on the Free plan—can use GPT-5 for general tasks, but higher-tier plans such as Pro and Team unlock “GPT-5 Pro” for extended reasoning sessions and higher usage limits. This tiered structure means that while basic capabilities are broadly accessible, heavy or specialized usage is better supported on paid plans. In a workplace context, ChatGPT Team and Enterprise plans also provide additional administrative controls, higher rate limits, and integration with internal tools.
In the API, GPT-5 is offered in several variants—gpt-5, gpt-5-mini, and gpt-5-nano—which differ in speed, cost, and reasoning strength. gpt-5 delivers the strongest performance and deepest reasoning, gpt-5-mini balances capability with lower cost and latency, and gpt-5-nano is optimized for the most efficient, high-throughput workloads. Pricing is usage-based, calculated per input and output token, with rates varying by model size. This lets developers choose the most cost-effective variant for their use case—for example, using gpt-5-nano for quick classification and gpt-5 for complex problem-solving.
Availability extends beyond OpenAI’s own platform. GPT-5 is accessible through Microsoft Azure AI Foundry, which allows integration into enterprise applications with Azure’s scalability and compliance features. In all deployment options, developers can also take advantage of parameters like reasoning_effort and verbosity to optimize cost versus performance. Because OpenAI has unified its model lineup under GPT-5’s architecture, developers benefit from consistent multimodal support and reasoning capabilities across different model sizes and platforms.