inclusionAI has released Ring-1T-2.5, a large language model (LLM) that promises state-of-the-art performance in "deep thinking".

Availability

The model is accessible via Hugging Face in FP8 format. This quantization level may allow for more efficient inference on hardware with limited computational capabilities, potentially making it suitable for on-premise or edge computing scenarios.

For those evaluating on-premise deployments, there are trade-offs to consider carefully. AI-RADAR offers analytical frameworks on /llm-onpremise to evaluate these aspects.