Follow the latest coverage, related explainers and connected technology stories.
Subquadratic announced the SubQ Snap decision model, which, according to the company’s tests, can process up to 384,000 tokens in a single request, while claiming to outperform Jev in most published tests. The company says the model is intended to reduce AI agents’ reliance on large language models for selection and routing tasks.
Microsoft launched Microsoft-Decision-1 for decision-making in applications, agents, and workflows after training it on Alibaba’s Qwen3.5-9B. The model is available at $0.042 per million input tokens with free outputs, but it currently lacks image support and open weights.
Cloudflare announced Clef and Clef-flash for making structured decisions and classifications with specified probabilities, hosting them on Workers AI and releasing them as open source under the Apache 2.0 license. It also unveiled a reinforcement-learning fine-tuning service that enables customers to customize the models for their data and use cases.