Cloudflare Clef: AI Model That Just Makes Decisions|flux ai news
Cloudflare launched Clef and Clef-flash, open-source AI models designed to make fast decisions instead of generating long text.
Source: https://developers.cloudflare.com/changelog/product/workers-ai/

Cloudflare has introduced Clef and Clef-flash, a new type of AI model designed to make decisions rather than generate long-form conversations. The models are available through Cloudflare Workers AI and are aimed at applications where an AI system needs to choose between predefined actions quickly.
Traditional large language models are often used for many different tasks, including writing text, answering questions and reasoning through complex problems. However, not every step in an AI application requires a model to generate a detailed response. Many workflows simply need a decision such as whether a request should be routed to sales, sent to customer support, blocked or escalated to a human.
This is where decision models such as Clef can be useful.
A simple agent workflow could look like:
Customer message → Clef → Decision → Automation
For example, imagine an online store receiving thousands of customer messages. Instead of using a large language model to write a complete response for every request, a decision model could classify the request and select an appropriate action.
A customer asking about a refund could be routed to support. A potential sales inquiry could be sent to a sales agent. A suspicious request could be blocked or escalated for human review.
The key difference is that Clef is designed to work with a predefined set of choices. Rather than producing paragraphs of text, the model can evaluate the available options and return probabilities associated with those choices.
This approach can be valuable for AI-agent systems because it separates decision-making from execution. One smaller model can decide what should happen, while another system or agent performs the actual task.
For example:
Ticket arrives → Clef evaluates it → “Support” selected → Support agent handles the ticket.
This architecture can potentially reduce unnecessary use of large language models. If an application only needs to choose between several known actions, generating a long explanation may add latency and cost without providing much additional value.
Cloudflare has also made the Clef models open source under the Apache 2.0 license, giving developers more flexibility to experiment with the technology and integrate decision models into their own applications.
The broader lesson is important for the future of AI agents: not every component of an intelligent workflow needs to be a giant language model.
A sophisticated agent system could combine several specialized components—one model for understanding a request, a decision model for selecting the correct action and another agent for executing that action.
This could lead to AI systems that are faster, more efficient and easier to control.
The shift is therefore not simply about creating another AI chatbot. It is about building AI systems where different models perform different jobs.
AI doesn't always need to talk. Sometimes, it just needs to decide.