open weights

Open-weight safety classifiers for inputs/outputs; tune to your policies.

Hugging Face repo for DBRX Instruct checkpoints under an open license for local inference and finetuning.

Ultra-small open models (135M/360M/1.7B) tailored for on-device and constrained local deployments.

Open multimodal family (1B–78B); 78B surpasses 70% on MMMU; broad image/video understanding.

Open-weight 32B RL-trained reasoning model reported to rival larger systems while remaining locally runnable.

01.AI’s upgraded Yi family (6B/34B) with improved instruction following and coding; widely used in local stacks.

Microsoft’s 14B open-weight reasoning model yielding strong complex-task performance with modest hardware.

Google’s open-weight 9B/27B models; practical to run locally and widely adopted for fine-tuning and apps.

Latest Qwen series from 0.5B–72B; dense decoder models with strong general, coding and math variants for local use.

Family of open-weight models (1B–90B, some with vision) designed for local and edge deployment and strong instruction following.