Make Adds OpenAI GPT-5.4 Mini and Nano Models for Faster, Cost-Efficient Automation.

Make has added support for OpenAI's newly released GPT-5.4 mini and GPT-5.4 nano models, giving users access to faster, more cost-efficient AI processing for high-volume automation workflows.

The two models target different use cases within Make's automation platform. GPT-5.4 mini delivers near-flagship reasoning capabilities at roughly twice the speed of previous models, while GPT-5.4 nano is positioned as the most cost-efficient option for repetitive, high-volume tasks. Both models are now available through the Make AI provider, which requires no separate OpenAI account, or users can connect their own OpenAI API keys.

The new models are accessible in three places within Make. First, they're available in the OpenAI app modules: Simple Text Prompt, Generate a Response, and Generate a Completion. These modules allow users to build scenarios that call OpenAI's API directly.

Second, the models integrate with Make's AI Agents framework, where mini and nano can function as subagents in more complex agentic workflows. This matters for organizations designing multi-step automation that requires different AI models at different stages, such as using nano for initial data processing and mini for more nuanced decision-making.

The rollout follows Make's broader push into agentic automation, including recent updates like MCP Toolboxes for controlled agent access and an agent library with pre-built templates.

The introduction of these models matters most for users running automation at scale. GPT-5.4 nano's cost efficiency makes it viable for tasks like bulk data classification, content moderation, or simple customer support triage where flagship models would be overkill. GPT-5.4 mini sits in the middle, offering stronger reasoning than nano without the latency or cost of flagship models like GPT-4.

Make's decision to offer these models through its own AI provider simplifies onboarding. Users can test and deploy scenarios without managing separate API credentials, though the option to bring your own key remains for those who need direct billing control or have existing OpenAI credits.

For teams already using Zapier or n8n, the update puts Make on similar footing in terms of AI model access. The ability to mix models within a single scenario, using nano for high-volume steps and mini for nuanced ones, gives Make users flexibility in optimizing both performance and costs across complex workflows.

The models are available now to all Make users through the OpenAI app and AI Agents modules.