Contrastive-LM has released CLM-8B, an open-source model that scores candidate actions for AI agents up to 9 times faste...

Contrastive-LM has released CLM-8B, an open-source model that scores candidate actions for AI agents up to 9 times faster than TypeSafe’s proprietary Jev system. The model uses a Qwen3-8B encoder with trainable projection heads trained on 60 million question-answer pairs. It achieves 81.6% accuracy on DeepSWE coding tasks and 87.6% on Terminal-Bench, running on a single NVIDIA GPU. https://www.marktechpost.com/2026/09/23/contrastive-lm-releases-clm-8b-an-open-system-one-model-that-scores-agent-actions-up-to-9x-faster-than-jev/ #AIagent #AI #GenAI #AgenticAI