Do you support multimodal models?

All models are currently text only. This is on our roadmap, no ETA yet. Feel free to reach out to your Cerebras point of contact with feedback and feature requests!